1869. ] DR. J. E. GRAY ON THE TORTOISES. 175 
in the British Museum, which have hitherto been referred to the 
family Emydide as defined in the ‘ Catalogue of Shield Reptiles in 
the British Museum,’ published in 1855, I think they may be more 
conveniently divided into four very natural groups, which may be 
called families. These groups may be thus characterized :-— 
1. The temporal muscle covered with skin, and generally protected 
by a narrow band-like zygomatic arch. 
1. Cistupinip#. The sternum united to the thorax by a cartila- 
ginous lateral suture, and divided transversely into two move- 
able portions. 
2. CueLyprRApD#&. The sternum united to the thorax by a bony 
symphysis, covered with from 7 to 11 shields; the middle por- 
tion fixed to the thorax; the front and hinder portions often 
separated from it by a transverse suture and moveable. 
3. Emypip#. The sternum united to the thorax by a bony sym- 
physis, solid, and covered with 12 shields. 
11. The temporal muscle covered with a bony hood formed by the ex- 
tension of the zygomatic arch. Head very large. Sternal 
shields 11. 
4, PLATYSTERNID&. Asiatic. 
Fam. I. Cisrup1n1p or Box-Torro!sss. 
Head moderate, covered with a hard thin skin. Eyes lateral or 
subsuperior ; pupil annular. Temporal muscle covered with the 
skin and (except in Cistudo) protected by a band-like zygomatic 
arch. Thorax covered with horny plates. Sternum very broad, 
attached to the thorax by a ligamentous suture, covered at the sides 
by the pectoral and abdominal shields, and divided across into two 
parts by a suture between the pectoral and abdominal plates. Sternal 
shields 12; the axillary and inguinal plates very small or wanting. 
The mastoid bone is excavated to form a tympanic cell. 
I have little to add to my monograph of the species of the family 
printed in the ‘ Proceedings’ of the Society for 1863, p. 173, except 
that the temporal muscle of the North-American genus Cistudo is 
only covered with skin, and the skull is destitute of any zygomatic 
arch between the orbit and the tympanic bone. In this respect, as 
well as in the position of the suture between the sternum and the 
thorax, this genus differs from the Lutremys of Europe and the 
genera found in Asia, all of which have a well-developed zygomatic 
arch for the protection of the temporal muscle. 
The skull of Lutremys of Europe is figured by Cuvier, Bojanus, 
and Wagler. Iam not aware that the skull of the very common 
Cistudo clausa has been figured or described. I have not seen any 
specimen of the Californian Cistudo blondinsia ; but, judging from 
the figure of the animal in Holbrook’s ‘ North-American Herpeto- 
