972 17. KINOSTERNIDE—18. TESTUDINIDZ 
on the average, much less mottled, but some specimens from 
Arizona show little difference in this respect. The difference 
in shape of the ninth marginal plate is said to be constant. I 
believe the two forms may be found to intergrade, and will 
be regarded as eastern and western subspecies. However, 
since this has not been clearly shown, and probably cannot 
be until much more material has been secured, I conform 
to the current usage and record two species from Arizona. 
Distribution —Dr. Stejneger informs me that, in addi- 
tion to the cotype said to have been secured in the Gila River 
at Camp Yuma, the National Museum has specimens of 
K. flavescens collected by Dr. Mearns and Col. Carpenter 
at Fort Verde, and by Mr. Julius Hurter in Graham County, 
Arizona. 
The National Museum has a specimen said to have been 
collected in Utah. 
Family 18. TESTUDINID/© 
This widely distributed family contains a large number 
of turtles distinguished from others chiefly by osteological 
characters. The shell is firmly ossified, and covered with 
large horny plates of which 11 or 12 are on the plastron. 
The pectoral plates are in contact with the marginals. The 
latter are 24 or 25 in number. The nec’ -an be completely 
drawn into the shell. Five genera are represented in the 
area under consideration. 
Synopsis OF GENERA 
a.—Plastron not hinged, not closing against carapace to form 
a tight box. 
b. 
Feet not club-shaped, webbed; two supracaudal plates; 
skin on top of head not divided into scales. 
c.—-Suture between abdominal plates less than twice 
length of suture between pectorals. 
