The Soft-shelled Turtles 



ties." The title is a harsh one, but it conveys the right 

 impression of their structure, as removed from the water 

 they present a ludicrous impression of large and ani- 

 mated pancakes. Their progress under such conditions is 

 clumsy and the soft plastron becomes bruised and lacerated on 

 a hard, rough surface. 



A key to the North American species is given :* 



A. Front margin of carapace with conical tubercles. 



Carapace brown or olive, with dull blotches or 

 black rings. Two pale bands on bead, forking a short 

 distance in front of eyes. 



SOUTHERN SOFT-SHELLED TURTLE, T. ferOX. 



Distribution. Georgia to Florida; westward to 



Louisiana. 



Carapace olive, with numerous black rings. 

 Two pale bands on bead, forking at base of proboscis. 



SPINY SOFT-SHELLED TURTLE, T. Spinifer. 



Distribution. Central and northern tributaries of 

 the Mississippi and St. Lawrence Rivers. 



B. Front margin of carapace smooth. 



Carapace brown or olive, unicolour or with obscure, 

 line-like blotches. Head markings obscure, but 

 forking at base of proboscis. 



BROWN SOFT-SHELLED TURTLE, T. muticuS. 



Distribution. Similar to T. spinifer. 

 Carapace olive, usually with numerous black dots. 

 Head bands uniting a short distance in front of eyes. 

 EMORY'S SOFT-SHELLED TURTLE, T. emoryi. 



Distribution. Tributaries of the Rio Grande River. 



The markings on the carapace are usually lacking on old 

 individuals of all the species. 



The descriptions of these turtles, when dealing with struc- 

 tural characters alone are difficult problems for the beginner. 

 The colouration of the carapace cannot be relied upon owing 

 to the fading of the markings on old specimens, as explained, 



"* Two alleged species are excluded. These are Trionyx asper and T. 

 nuchalis, both of Agassiz. The range of the former has been given as 

 : 'The Mississippi" the latter as occurring in the Cumberland and the 

 Tennessee Rivers. The writer has been unable to examine specimens 

 of either and is not convinced as to their standing as full species. Both 

 appear to be related to T. mutica. 



73 



