688 DISCONTINUOUS DISTKIBUTION OF MA,MMALS. 



does. Lydekker says : " Some naturalists separate the African 

 genera from the Octodontidse as a separate family, Ctenodacty- 

 iidae. Such an arrangement, however, tends to obscure the inti- 

 mate connection existing between the African and American 

 forms, and is therefore undesirable." Beddard still includes one 

 African genus, Pteromys, in the Octodontidae (American genera), 

 but says that it should be, rather, as it is by Thomas, 

 referred to the Ctenodactylidas, which would clear up 

 the geographical anomaly. This is again a case of discontinuous 

 distribution, such as that of the Lemurs, Marsupials, and 

 Primates, and also here we find the group showing archaic 

 characters ; such are, firstly, the teeth, which in Pteromys are 

 described thus by Beddard : " Their surface is uneven, and differs 

 from other Hystricomorphs in that the enamel to the inside 

 of each upper jaw-tooth and outside on each lower jaw-tooth 

 forms two tubercles " ; secondly, in that a prehensile tail is still 

 retained by the Cuban Hutia (Capromys prehensilis). 



That these very similar animals have been divided according 

 to their geographical occurrence is, of course, a tacit acceptance 

 of the hypothesis that they are phyletically distinct, but have fol- 

 lowed parallel lines in their evolution. 



Further, the above families of Hystricomorphs more or less 

 lead over to the Porcupines ; and the singular fact in this is that 

 the Octodontidae merge into the American Tree Porcupines (Cer- 

 colabidas), whereas the Ctenodactylidas must be connected to the 

 Old-World Porcupines (Hystricid^e). 



Now these animals have tails of variable length, and the 

 American forms have still the Creodont prehensile tail ; we can 

 illustrate their possible ancestry thus : — 



A. Creodont ancestor in America 



(also ancestor of the Marsupials?) 



ITutia-like ancestor 



Tree porcupines. Octodontidse. Cricetidae. 



(white- footed mice 

 of America) 

 B. Creodont ancestor in Africa 



y (also ancestor of Lemurs?) 



Pectinator-like form 

 (somewhat Squirrel-like )s. 



'^^ Ctenodactylidae. 



Aulacodus-like ancestor. 



I , - . 



Porcupines. 



