THE UPPEE SILUKIAN OF NORTH AMEEICA. 53 



tion in thickness above mentioned is produced. In almost all 

 specimens the upper surface shows adhering reddish or yellowish 

 matter, from which it is difficult to free it. This matter is doubtless 

 a portion of the crystalline substance filling the cells of the next layer. 



Some have used the term " enamel " in writing of this layer, but 

 I have been quite unable to see any trace of that structure in it. 

 Instead of prisms it rather appears to be stratified, so that the term 

 " nacreous," applied to Ptemsjns by Prof. Huxley, would more accu- 

 rately describe it. There is, however, no appearance of the pearly 

 lustre that characterizes true nacre. 



The second layer (6) forms the thickest portion of the shield, and 

 may be termed the cellular layer. It consists of a number of cells 

 formed by party walls of a material resembling in appearance that 

 composing the basal layer. These cells have no constant outline, 

 but for the most part have 4, 5, or 6 sides. Nor have they one size, 

 the larger being as much as -j^ of an inch in their longest dimensions, 

 while others measure less than a fourth of this. In shape too they 

 range from nearly square to oblong and hexagonal. The w^alls that 

 form them are equally variable in size, ranging from -^-q to -j-i-^j- of 

 an inch in thickness. The angles of these cells are rounded, the 

 partitions being thicker there than elsewhere. The walls stand for 

 the most part vertically, or nearly so, on the basal layer, and sup- 

 port on their other ends the third layer next to be described. As 

 already mentioned, some of the circidar vessels enter them, and re- 

 main in them for some distance, perhaps through their whole 

 height ; others lie partially in the cell-wall, and partially in the cell. 

 The height of the walls, and consequently the thickness of the second 

 layer of the shield is about five times that of the basal layer, or about 

 ^ of an inch. The cells are filled with infiltrated calcareous matter, 

 which under the action of the weather is dissolved out, leaving 

 an exceedingly brittle cellular mass to represent the original shield. 



The third layer (c) deserves the name of the vascular layer, con- 

 taining as it does a perfect system of vessels imbedded in its sub- 

 stance. This substance is the same in appearance as that which 

 composes the basal and cellular layers, and is continuous with them. 

 Instead, however, of being solid, or nearly so, it is traversed by 

 vessels, the presence of which enables us to divide it into two por- 

 tions. The first and lowest of these lies immediately over the 

 covering layer of the cells, and contains a set of largo vessels running 

 parallel with the striation of the outer surface (fig. 4). The diameter 

 of these vessels is equal to the thickness of the covering and basal 

 layers, or about g-J-jj of an inch. They are perfectly even in outline, 

 but retain approximately the same diameter, as far as I have traced 

 them. These vessels communicate with the cells through openings 

 in their covering layer, apparently forming in this way a connexion 

 with every cell over which they pass. 



They likewise communicate by short branches of about half their 

 own diameter, and tapering toward their outer ends, with another 

 set of vessels or channels running parallel with and between them, 

 of somewhat smaller calibre and more regular margin, lying slightly 



