634 ERNEST WARREN. 



coluTiinai', aud would correspond to the enamel epithelium 

 of an ordinary tooth. 



In the case of the ctenoid lingual teeth the tooth-germs 

 have fused together, and the cones of large granular cells, 

 which will be converted into the succeeding horny tooth, 

 are continuous with one another, although small separate 

 pulp-cavities can be distinguished to each cone or originally 

 distinct tooth-germ. 



The structures described do not lend support to the idea 

 that these horny teeth are degenerate calcified teeth, but if 

 tliey actually are degenerate they must be regarded as having 

 reverted to a condition that probably preceded the placoid 

 scale of an Elasraobranch. 



The placoid scale is of very great antiquity, being found 

 amongst the oldest organic remains, and living and fossil 

 forms give us little information as to its origin. The stages 

 in the evolution of a placoid scale must, therefore, be sur- 

 mised rather than actually observed. 



A rough skin was undoubtedly of some use to the ancestors 

 of Elasrabbranchs, and the simplest condition conceivable 

 would be where the general surface of the body and jaws 

 was covered with little horny warts. As the warts were 

 gradually converted into more pronounced structures they 

 Avould come to possess a pulp-cavity, in response to the need 

 of a supply of blood-vessels, etc., to the proliferating epi- 

 dermis. Such specialised warts where cornification was very 

 complete would result in structures like the horny teeth of 

 Petromyzon and Myxine. 



Calcification of the outer portion of the pulp would add 

 strength to the horny scale, and there is no improbability 

 against supposing that such a variation arose. The shape of 

 the resulting calcified mass would be moulded by the over- 

 lying Malpighian layer of the skin (enamel epithelium), just as 

 the shape of the future tooth is prefigured by the enamel 

 organ. 



If the horny teeth or warts were useful to their possessors, 

 then a projecting cone of calcified substance would certainly 



