N. ROSÉN, STUDIES ON THE PLECTOGNATHS. 9 



cations are spineless. They have the form of small plates 

 and represent a somewhat early stage in the evolution of 

 the integumental ossifications (Fig. B). Their place is in the 

 deepest part of the exteriör layer of the corium. That they 

 are formed by cell-action seems to be quite certain from the 

 fact, that their surface is coated with cells. Their structure 

 is homogeneous. Another fäet concerning the earliest forma- 

 tion of the plates is to be noted. The epidermis runs quite 

 straight, forming no papillae, from which the plates originate. 

 When we compare the early stages of the integumental ossi- 

 fications just described with their structure in the fully 

 developed form, I think their course of evolution is quite 

 evident. The small plate first formed grows in the beginning 

 in all directions owing to cell-action (Fig. C, a). As soon as 

 its lower surface has reached the fibrillated layer of the 

 corium, another way of formation takes rise parallel to the 

 cellulär action. The fibres of this layer of the corium which 

 run along the plates increase considerably in number, become 

 connected more closely to each other. Their course becomes 

 very winding and the tissue formed by them calcified (Fig. 

 B and C). In what way the calcification takes place is dif- 

 ficult to say, but it is, I think, most probable, that it is due 

 to the action of the cells in the fibrillated layer. By this 

 calcification of a part of the fibrillated layer of the corium 

 another stratum is added to the homogeneous one first form- 

 ed. In the fibrillated part of the plate I have not been 

 able to find any cells. The homogeneous part of the ossifi- 

 cations grows chiefly in a direction towards the epidermis, 

 thus forming the spines. The basal part of the spine is 

 formed long before the fibrillated portion of the plate has 

 taken its origin. That the growth of the spine takes place 

 by the addition of new substance of homogeneous structure 

 on the surface is evident. The spine is solid, a growth from 

 a pulp being impossible. As already mentioned the growth 

 of the spine is by no means uniform. There must be periods 

 of intensive growth alternating with slower growth. A trans- 

 verse section through a spine shows clearly such different 

 zones of growth (Fig. 5). The relation of the epidermis to 

 the development of tbe integumental ossifications will best 

 be understood by studying Fig. A and B. As is shown in 

 the diagram of a section through the skin of a fullgrown 



