336 EMBRYOLOGY OF THE LOWER VERTEBRATES en. 



only proceed with the process when room is provided by the already 

 developed scales becoming spaced out during the growth of the 

 body. 



While the placoid scale is simply an individual dermal tooth 

 the ganoid scales as seen in the surviving Poly pier us or Lepidosteus 

 are on the other hand tooth-plates, numerous minute denticles being 

 associated with each scale. In these fishes also there is a certain 

 amount of independence between the dermal plate of bone and the 

 actual denticles which are at first quite separate from it (Nickerson, 

 1893 ; Goodrich, 1908). The reduction in size of the dental cones 

 and the loss of their attachment to the bony plate are steps towards 

 their complete disappearance which has been reached in the scales of 

 ordinary Teleostean fishes. 



In the ganoid scale of Polypterus or Lepidosteus the protective 

 power of the bony plate has been greatly increased by its superficial 

 layers undergoing modification of an analogous kind to that of the 

 superficial layers of the dentine cone in the teeth of fishes. This 

 portion of the scale is extremely dense, hard and enamel-like and is 

 without cells embedded in it. Like the corresponding layer in the 

 tooth of a fish it is commonly known by Williamson's name Ganoine. 

 The advisability of using this name, rather than enamel, rests mainly 

 upon the assumption that enamel is a substance fundamentally 

 different, derived from a different cell-layer, from bone or dentine. If 

 it be the case however that enamel is merely the superficial layer 

 of dentine which has undergone secondary modification then there 

 seems no particular harm in adhering to the custom until recent 

 years quite general of using the word enamel for the superficial 

 layer of the ganoid scale. 



Ganoid scales are still comparatively thick and bulky struc- 

 tures but in the typical Teleosts the scales have become very thin 

 plates of bone so modified as to be very tough and flexible, and 

 overlapping like slates on a roof so as to be able to slide over one 

 another during flexure of the body. This overlapping has been 

 rendered possible owing to the surface of the scale being no longer 

 inseparably linked to the ectoderm by the development of teeth. 

 The scale is developed in the thickness of the dermis and it is only 

 at its posterior edge, if anywhere, that it is even in early si 

 connected with the epidermis. 



Vestiges of dermal denticles have been described in Teleostean 

 fishes and these deserve fuller investigation. Marett Tims (L906) 

 describes a sta^e in f,W//.s in \\hieh tin- sc.;i.le consists"!' separate 

 platele! each \\iili a tooth-like spine projecting from it, while in 

 such South American Siluroids as CW//V///7///N ami Lric<t ,i<i tin- 

 plates of the bony cuirasse bear numerous small spines \\liicli appear 

 to be 1 vpieally tooth-like in structure. 



Tim scale is a plate <!' hone immersed in the dermis and it 

 then-ion- naturally -rows by the addition of new U.nc all over iN 

 In the ganoid scale the ijuality <>!' the l.one diller- on t he 



