No. 384.] CHELONIAN CARAPACE AND PLASTRON. 947 
have corresponding gastralia. There could then not have been 
more gastralia than would correspond to two pairs of ribs. 
Excluding the entoplastron and epiplastra, which originated 
otherwise, we have in front of the umbilicus, in most turtles, a 
single pair of plastral elements, the hyoplastra. But it is evi- 
dent that in the earliest turtles there was an additional pair, the 
mesoplastrals. They are present in some living Pleurodira. In 
the genus Sternothzrus! the mesoplastrals extend right across 
the plastron, and meet in the mid-line. In Pelomedusa? and 
Podocnemis è they are reduced to the condition of wedge- 
shaped plates lying on the bridge. In all the Cryptodira and 
the Trionychia these plates have been extirpated. If, then, we 
have not assigned too many pairs of ribs to the sternal region, 
there was originally involved in the preumbilical region the 
gastralia belonging to only two pairs of ribs, those of a pair of 
ribs for each pair of plastral elements. 
It is possible that six sets of gastralia entered into the com- 
position of the hypoplastra and xiphiplastra. However, we do 
not find, in Sphenodon or any other forms, that the pubic 
region is covered with gastralia. In Sphenodon there are three 
lumbar vertebrae, and some of the hindermost gastralia are 
much reduced in extent. If the plastron of most turtles ex- 
tends beneath the pelvic region and even behind it, this is due 
doubtless to secondary modifications. The condition of the 
plastron of the Chelydridz is probably more primitive. We 
must therefore believe that some of the hindermost gastralia 
become aborted. I am inclined to the opinion that we have 
at present in the hinder portion of the plastron elements rep- 
resenting only two pairs of ribs. If the hyoplastra and the 
mesoplastra each were developed from the gastralia belonging 
to a single pair of ribs, the same thing appears probable in the 
case of the hinder plastral elements. Otherwise we must 
assume that there has been codssification of originally distinct 
bones ; but the manner in which the mesoplastra have been 
thrust out of the plastron in most turtles indicates that here 
1 Boulenger. . Cat. of Chelonians, p. 193, Fig. 47. 
. 2 OP. cit., p. 199, Fig. 49. 
8 OP. cit., p. 201, Fig. 51. 
