N. ROSEN, STUDIES ON THE PLECTOGNATHS. 7 



nued from the outher surface of one tubercle to that of the 

 adjacent ones, över the serrated apposed margins. In it, 

 and in it only, the pigment of the skin was contained.» 

 From these lines and from the figure given by Turner it 

 is quite evident, that he has committed a grave error about 

 the superficial layer of the pla*tes. The layer of which he 

 speaks forms no part of the plates nor of the spines of the 

 plates, but is the exteriör layer of the corium described 

 above. By studying microscopical slides of the plates as 

 well as plates isolated from the cutis I have found tvvo 

 distinct layers: one superior forming the spine and a thin 

 cover of the plates and one inferior making up the greatest 

 part of the plate (Fig. A and 4). As regards structure, these 

 two parts are rather different. The superior has quite a 

 homogeneous appearance, the inferior being distinctly fibril- 

 lated. This structural difference will best be understood, 

 when I say, anticipating the results of the study of the de- 

 velopment of the ossifications, that the homogeneous part 

 is formed by the secretion of the cells of the exteriör layer 

 of the corium, and that the fibrillated part is to be consi- 

 dered as a portion of the interiör layer of the corium to a 

 great extent densified and calcified. The bundles of fibres 

 anastomose with each other and run close to each other. It 

 is only in thin slides that it is possible to see, that the struc- 

 ture of the lower part of the plate really is fibrillated (Fig. 4). 

 The upper layer is not perfectly homogeneous. In horizontal 

 slides of the spines which have been decalcified and stained 

 with the mixture of van Gieson distinct darker lines can be 

 observed (Fig. 5). In what structural or chemical respect 

 these differ from the broad light belts of the spine I do not 

 know, but it seems to me that it is out of the question, that 

 these different layers of the spine represent zones of growth 

 in the same way as the belts of ordinary fish scales and 

 otoliths do. The upper surface of the plate is not smooth. 

 Fine furrows and lines radiating from the spine are to be 

 seen (Fig. 3). The lower surface of the plate is not straight 

 but concave. The plates are very often perforated forming 

 openings for the bloodvessels of the corium, which pass from 

 one layer to another. The plates are closely set with their 

 serrated margins locking into each other. The fibrillated 

 portions of the plates are connected with each other by 



