288 



TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



Tetraodonts. — It is noted above that in the balistids there are many sets of teeth; 

 experience with other groups, such as the elasmobranchs, the scarlds, oplegnathids, etc., 

 shows that a beak may be produced in such cases by the multiplication of these teeth and 

 the subsequent close appression or perhaps complete fusion. In the trunk-fishes (Ostra- 

 ciontids), as above noted, the teeth are. small and numerous and close-set but still distinct. 

 In the tetraodonts or puffers (e.g., Spheroides, Lagocephalus, Figs. 167-169) the beak is 

 complete except that the right and left halves remain quite distinct. Here the beak has 

 more of a shearing character, while in the porcupine-fishes (Diodontidae) the beak (Figs. 

 170, 171) is more massive and more adapted for crushing. 



g Lactophrys tricornis 



Fig. 166. Lactophrys tricornis. 



With the acquisition of sharp-cutting edges on the four dental plates of the puffers 

 there has probably been a rapid increase in the size of the mouth and a change perhaps 

 from nibbling to shearing; also the form of the body when uninflated is of a swifter type 

 than that of the other families, so that puffers ought to have a more omnivorous diet, 

 including crustaceans. It is therefore not surprising to read in Bigelow and Welch (1925, 

 p. 299) that puffers ". . . feed on small crustaceans of all sorts, especially crabs, shrimp, 

 and amphipods, as well as on small mollusks, worms, barnacles, sea urchins, and other 

 invertebrates, which they find on the bottom. Young fry of 7 to 10 mm., examined by 

 Doctor Linton at Woods Hole, had eaten copepods and crustacean and molluscan larvae." 



The general hypothesis that I am defending, namely, that the whole plectognath stock 



