1871.] MR. ST. GEORGE MIVART ON HEMICENTETES. 65 



The skeleton of Hemicentetes closely resembles that of Centetes, 

 except that the neural spines, especially the cervical ones, are rela- 

 tively, as well as absolutely, less developed, and that the dorso- 

 lumbar vertebrae are twenty, or at most twenty-one, in number, 

 instead of twenty-three or twenty-four as in Centetes. The pubic 



Fig. 9. 



Pelvis, once and a half the size of nature. 



symphysis is also widely open in some individuals (probably females) ; 

 and the humerus is not quite so long as the scapula. Moreover 

 the os scaphoides is distinct from the os lunare ; and there is no os 

 intermedium. 



Since the publication of my paper on the osteology of the Insec- 

 tivora* additional material has come to hand. Thus skeletons not 

 only of Hemicentetes (formerly known as Centetes madagasca- 

 riensis), but also new ones of Rhynchocyon and Petrodromus, have 

 been added to the British Museum and the collection of the Royal 

 College of Surgeons. At the last-named institution there has also 

 been received a perfect skeleton of Ericulas, which is here figured 

 by the kind permission of the Council and Curator. 



Rhynchocyon. — As to this genus I am now able to add that the 

 occipital foramen looks mainly backward, that the pterygoid fossa 

 does not nearly extend so far forward as the hinder margin of the 

 palate, that there is no paroccipital process, and that there is a very 

 small mastoid process just behind the external auditory meatus, but 

 a more marked projection at the lower end of the mastoid where it 

 runs down behind the auditory bulla at the posterior end of the 

 harmonia joining the tympanic to the petrosal. 



I may also add that, in a skull in the British Museum, I find there 

 are two minute teeth in the place of the first upper premolar. As 



* See 'Cambridge Journal of Anatomy,' vol. i. (1867) p. 281, and vol. ii. 

 (1869) p. 117. 



Proc. Zool. Soc— -1871, No. V. 



