444 MR. R. I, POCOCK ON THE 



in front of the /ora7nen ovale. When this aperture is absent, the 

 presence of the canal may be demonstrated by cutting away its 

 outer wall backwards from its anterior orifice. The true forame^i 

 rotundmn will then be revealed perforating the cranial floor at 

 the posterior end of the canal a little in advance of the foramen 

 ovale on the admedian side (text-fig. 1 , F). Hence in Viverricula 

 i\\e foramen rotundum opens into the alisphenoid canal, or, if the 

 term ' canal ' be inadmissible for a tube closed at one end, into 

 the alisphenoid tube representing the canal. 



" It may be added that this region of the skull in Viverricula 

 is alike in all specimens, apart from the presence or absence of 

 the posterior orifice of the canal, which may be represented by a 

 hole only large enough to insert a needle. Whereas if the 

 alleged absence of the canal were due to the suppression of its 

 outer wall, the canal would be represented by a groove, as in 

 Cynogale and Utcpleres, which is not the case, and the foramen 

 rotimdum in Viverricula would have to be described as a long 

 tube, to which no parallel can be found in the ^luroidea. 



" That the interpretation above given is correct may be f urthe':- 

 shown by comparing Viverricula with Genetta (text-fig. 1, G, H), 

 Civettictis, and other genera where the foraonen rotundum, piercing 

 the skull, may be seen within the alisphenoid canal by looking 

 through its posterior orifice, the aperture in the skull close to the 

 sphenoidal fissure being the anterior orifice of the canal and not 

 the foramen rotundum of the Felidfe. 



'' The alisphenoid canal is also stated in current literature to be 

 absent in the Hytenidfe ; and this opinion seems to date from 

 Turner's rejection in 1848 (P. Z. S. 1848, p. 81) of Cuvier's 

 statement in 1837 that it is present in these animals. Cuvier's 

 words are : — ' Dans I'hyene . . . . le trou optique, le spheno- 

 orbitaire, le rond, le vidien [alisphenoid caiial] et I'ovale difi"erent 

 pen du chien. J'ai un individu oii il y a un canal vidien d'un 

 cote et pas de I'autre ' (Anat. Comp. ed. 2, ii. p. 471). This 

 assertion, suggesting that the canal is generally present and 

 exceptionally absent, is not altogether correct ; nevertheless. 

 Turner, Flower, and Mivart were wrong in citing the absence of 

 the canal as characteristic of the family Hyfenidse. It is usually 

 absent but sometimes present, at all events in Crocuta (text-fig. 2). 

 It is much shorter than in Vxverricxda. Nevertheless, its apparent 

 absence is dvie to the same process as in that genus, namely the 

 obliteration of its posteiior orifice. There is sometimes no trace 

 of this orifice ; but quite commonly it is represented by a small 

 aperture a little in front of the foramen ovale. This aperture 

 may lead into a very short blindly ending tube, whence a small 

 hole, also to be seen a,t the posterior end of the canal in Canis, 

 penetrates the sphenoid bone (text-fig. 2, B). In other and rarer 

 cases where this aperture is larger, a bristle passed into it 

 emerges at a tolerably large foramen lying beneath the hinder 

 end of the sphenoidal fissure in the temjjoral fossa. This foramen 

 is the anterior end of the canal and not the foramen rotundum 

 which perforates the base of the skull within the canal behind 



