332 Proceedings of Royal Society of Edinburgh. [sess. 
In the crab it is found that the carapace is formed by a regular 
secreting membrane, which, according to Vitzou,* has the follow- 
ing structure. [This we have verified for ourselves, and give 
briefly.] 
There is, most externally, a thin delicate chitinous layer in which 
little or no structure can he made out. Beneath this is a thick 
chitinous layer, infiltrated with lime salts, which must be removed 
by maceration in weak acid before the structure can be dis- 
tinguished ; when so softened and examined in thin sections, it is 
found to he composed of a series of layers of chitin lying very 
regularly, one above another, this arrangement giving rise to a 
series of markings- more or less parallel to the surface. Running at 
right angles to these are other regular markings, the meaning of 
which we shall see immediately. 
These chitinous layers rest immediately on a layer of tall 
columnar cells, each with a nucleufe and a distinct nucleolus. If now 
we again examine the lines in the chitinous layer running at right 
angles to the surface, we find that the distance between them 
corresponds exactly with the breadth of one of these cells, so that 
each of these columnar cells may he said to secrete, time after time, 
its little area of chitin. In some cases the markings in the chitin 
running at right angles to the surface are more numerous than the 
intercellular spaces, in which case there is evidently a splitting, 
similar to that which takes place in the formation of the striated 
margin of a ciliated epithelial cell. 
These long columnar epithelial cells send down processes which 
rest on a very distinct basement membrane, and then beneath this 
basement membrane there is either a mass of muscular tissue or a 
layer of connective tissue, in which are large vacuolated-looking 
cells. The vacuoles contain a material which gives all the reactions 
of glycogen. 
It is evident from the above description that these columnar 
cells are the active agents in forming the chitinous covering of the 
crab, which really corresponds to the horny layer of the skin of an 
animal, or to the thick horny lining of the gizzard of a fowl. The 
only difference between these cells and those of the retce Malpighii 
being that one part of the cell is constantly growing outwards and 
* Loc. cit . , p. 501 et seq. 
