198 MOSTLY MAMMALS 



trace the ancestry. A German writer, the late Prof. L. 

 Fitzinger, considered that domesticated dogs might be 

 divided into seven well-marked groups, which included 

 close upon a couple of hundred of more or less well-marked 

 breeds and varieties. Other authorities are, however, of 

 opinion that the number of main groups might be reduced 

 to half a dozen, these including wolf-like dogs, such as the 

 Eskimo breed, the various kinds of greyhounds, spaniels, 

 hounds, mastiffs, and lastly terriers. 



All who have written on the subject are in accord in 

 regarding all domesticated dogs, with the exception of the 

 Australian dingo, as constituting but a single species — the 

 Cajiis familiaris of Linnaeus. But if it be true, as seems 

 probably the case, that domesticated dogs trace their 

 ancestry to more than a single wild species, it will be 

 obvious that Canis familiaris cannot in any sense be re- 

 garded as equivalent to an ordinary wild species ; and that, 

 properly speaking, if this were possible, the various true 

 breeds ought to be affiliated to the wild species from w^hich 

 they are respectively derived. Still, for practical purposes, 

 the ordinary classification may be accepted, if it be remem- 

 bered that Canis familiaris, like Felis domestica, is in all 

 probability a "convergent" species. 



By naturalists all the members of the dog tribe are in- 

 cluded in the great family Canidae, which thus embraces 

 wolves, jackals, foxes, wild dogs, the African hunting-dog, 

 the long-eared fox of the Cape, and the bush-dog of Guiana. 

 Somewhat different views are entertained as to how many 

 of these should be included in the typical genus Canis, but 

 this is a matter which needs no consideration here, and we 

 may accordingly proceed to eliminate from the list those 

 groups which have certainly no claim to be on the ances- 

 tral line of the domesticated breeds. 



