THE INTEGUMENT AND THE EXOSKELETON 85 



in character and never developing far enough to break through 

 the gums. 



The scales of mammals are commonly little emphasized, 

 owing to the conspicuous nature of the hairy coat, principally 

 associated with them, but they are, nevertheless, of great mor- 

 phological value. They occur in definite regions and only in 

 certain forms, but are so widely distributed that, were all 

 other reasons absent, their former more extensive distribu- 

 tion would be strongly suggested. In most cases they are 

 found only on tails and paws, but in the Manidce, an edentate 

 group, the entire dorsal surface of the body and limbs is cov- 

 ered with large, imbricate scales, and in the closely related 

 armadillos, similar scales fuse to form a dorsal carapace, as 

 well as shields for the head, tail and limbs. Scale formation 

 on paws and tail occurs mainly in marsupials, rodents and 

 insectivores, and may be seen particularly well on the dorsal 

 surface of the paws of moles and shrews, or on the flat tails 

 of the beaver and muskrat, in which the scales are usually 

 rounded and regularly imbricated. Where the tail is cylin- 

 drical, as in the rats and mice, the scales are arranged in rings, 

 those of one row standing in imbricated relation to the one 

 which it overlaps. 



In structure the scales are epidermic, like those of reptiles, 

 underlaid by corium papillae. They usually remain more or less 

 embryonic, and the epidermis, though cornified, does not de- 

 velop definite hard parts, but in the Manidce distinct horn 

 scales are produced, as thick and heavy as those of reptiles, 

 the main difference being that there are here no periodic 

 ecdyses, and the scales are shed and renewed singly, as occa- 

 sion requires. In young armadillos the scales that form the 

 carapace and shields are like those of Manis, but they become 

 soon reinforced by ossifications of the corium, one for each 

 scale, which enlarge and finally fuse to form an osseous sub- 

 structure. These corium elements are plainly secondary struc- 

 tures and are not to be considered as primary elements of the 

 mammalian scale, which, as stated above, is entirely epidermic. 



Aside from the sporadic occurrence of scaled areas in 



