ventral surfaces of the concrescence, in which case the\' are invariably lilled with 

 a substance now identified as a manganese compound. As endodermal cells 

 they may, therefore, be said to be not regularly well dehned as such in Calliti'is 

 leaves, and in this respect there is a resemblance to those of Sciadopitys of Japan. 

 The cell walls of the leaf tissue generally are irregularly circular in section, or 

 having a slight tendency to hexagonal form, and they show no in\-olutions or 

 infoldings, so characteristic of Conifer leaf cells in general. 



In the preparation of the sections, the protoplasmic contents of certain cells 

 ha\-e been remo\-ed, and so they in^•ariably appear empty, and it is thus that 

 they are easily differentiated from the tracheids of the transfusion tissue. The 

 mesophvll needs little comment. It consists of spongy and palisade parenchyma, 

 and both are clearly dehned in Figures 65 to 76. The latter consist of a single 

 row having the long axis at right angles to the dorsal surface of each leaf, but 

 cease at the ventral curve. The thick-walled hypodermal cells are, so to speak) 

 the epidermal cell companions of these, as they also only extend as far as the 

 epidermal and palisade cells, and gradually diminish in size and finally give out, 

 as they approach the ventral surface. They are largest and thickest walled at 

 the apex of the dorsal cur\'e, and generally number about 100. The epidermal 

 dorsal cells may be described as rectangular, and like the hypodermal ones are 

 largest at the dorsal apex where the outer cell wall or cuticle is much thickened. 

 They are not so numerous as the hypodermal cells, fifty being about the limit. 



The cells of the ventral surface take quite a different form from those of 

 the dorsal, for as they turn, so to speak, to curve into the ventral surface, the 

 thick cuticle walls gradually dome until in the centre of the ventral cavity of the 

 concrescence these walls reach their maximum height, becoming quite conical in 

 shape — the elongated apices appearing to resemble numerous cones. They 

 are clearly illustrated in Figure 76. This unusual structure, as far as we are 

 aware, has onlv been recorded in one other instance in Conifers, i.e., Sciadopitys 

 verticcllata , S. and Z. of Japan ,C. F. Bertrand, " The Gnetaccce et Coniferce," 



pi. X, Figs. 10, II, 12;. 



The function of these elongated bodies (jr pa])ill()sc projections is probably (i) 

 to assist the guard cells in the performance of tli(ir innction or duties, (2) they 

 also indicate the presence of the stomata, being only louiid along with them, 

 ; a protective character for the stomata by closing over tlieni as occasion 

 requires during adverse climatic or other conditions, and (4; eventually seed 

 protectors, for in the transitif)n of the terminal leaves into cone scales, these 

 elongated cells interlock with those on the opposite leaf Tsporophyll) like teeth 

 of a cogwheel, and becoming ligneous, hold tlic cells togetlur in a very firm grasp 

 during the maturing of the seeds (Figure 17). The guard cells of the stomata 

 call for little comment, being of the usual shape of such, relatively to tlie size 

 of the air cavities, and much sunk below the cuticle. 



