Fishes of the Western North Atlantic 251 



likely that the change in the specific composition of the crab fauna (on which they largely 

 subsist) from the waters west of Cape Cod to those to the east can be responsible. 



Present indications also are that this particular population is bounded equally sharply 

 to the southward, for while "Smooth Dogs" are common in season off the coasts of south- 

 ern Virginia'" and North Carolina, at least as far as Cape Lookout, there are only three 

 reports of the species for South Carolina (including one specimen in the collection of the 

 Harvard Museum of Comparative Zoology), and no positive record or rumor of its 

 presence on the east coast of Florida." 



In contrast with the considerable amount of information that has accumulated about 

 this species along the east coast of the United States, little is known of it in the Gulf of 

 Mexico and the Caribbean, except that it occurs on the coast of Texas,''^ around Cuba and 

 Jamaica in some numbers (p. 250), at Curagao,^' and at Trinidad.^* Whether the Cuban 

 center of population receives any recruits in winter from the northern stock is not known. 

 Positive knowledge of the distribution farther south is even more scant, for while it has 

 been reported repeatedly by name from southern Brazil, Uruguay and northern Argen- 

 tina, there is no knowing how many of these records actually may have referred to the 

 newly described schmitti (p. 261 ).^° But M. canis does occur in southern Brazil, as proved 

 by the fact that the collection in the Harvard Museum of Comparative Zoology includes 

 several small specimens from Rio de Janeiro, as well as 3 1 embryos and the head of a large 

 female from an indeterminate Brazilian locality that we cannot distinguish from canis. 

 And we judge from an excellent illustration"" that a Mustelus, said to be the most common 

 Uruguayan shark," is likewise identical with the North American canis. Comparison also 

 of a specimen from Bermuda with extensive series from North America, West Indies and 

 Brazil failed to reveal differences sufficient to warrant specific separation."* The coastwise 

 nature of this species makes it likely that the Bermudian population has long been entirely 

 isolated. 



Synonyms and References:"' 



S^Wtt/ <ra«!V Mitchill, Trans. Lit. Phil. Soc. N. Y., /, 1815:486 (type loc, N. York). 



Mustelus canis DeKay, Zool. N. Y., 4, 1842: 355, pi. 64, fig. 209 (descr., N. York) ; Linsley, Amer. J. Sci., 

 4y, 1844: 77 (Connecticut); Storer, Mem. Amer. Acad. Arts Sci., 2, 1846: 505 (in synopsis); Baird, 



20. Only occasional specimens are taken in Chesapeake Bay. 



21. Canis has been recorded from southern Florida, but it is probable that these reports actually referred to the 

 newly discovered M. norrisi, for which that is the center of abundance (p. 255). 



22. See Study Material, p. 244. It has also been reported as common in fresh water in Louisiana (Fowler, Proc. 

 biol. Soc. Wash., ^6, 1933 : 57), but perhaps not on good evidence. 



23. Measurements given by Metzelaar (Trop. Atlant. Visschen, 1919: 5) identify this specimen as canis, not norrisi. 



24. We have examined the embryo reported by Fowler (Proc. Acad. nat. Sci. Philad., 6y, 1916: 521) from Trinidad. 



25. For discussion of this species, see Bigelow and Schroeder (Proc. Boston Soc. nat. Hist., ^i [s], 1940: 423). 



26. Devincenzi and Barattini, An. Mus. Hist. nat. Montevideo, Suppl. Atlas, (2) 2, 1926: pi. 2, fig. 2. 



27. Devincenzi, An. Mus. Hist. nat. Montevideo, (2) /, 1920: 12. 



28. Bigelow and Schroeder, Proc. Boston Soc. nat. Hist., 41, 1940: 417. 



29. The many studies of the physiology of Mustelus canis are omitted here, unless pertinent to knowledge either 

 of itf geographic distribution or of its habits. 



