220 



ORTHOGENETIC VARIATION IN THE SHELLS OF CHELONIA. 



lateral pair became the costal, the outermost or most lateral the marginal set. The 

 intermediate row between these two still survives in some recent genera as the so- 

 called supramarginals ; it became gradually suppressed owing to the increasing size of 

 the costals. 



The last costals, say those of the 18th to 20th metameres, became likewise 

 suppressed, in conformity with the shaping of the trunk ; the three last neurals were 

 turned into pygals and the last pair of marginals closed round the posterior end, 

 meeting in the middle line, and there they occasionally fuse, for instance in T. mauri- 

 tanica, into an unpaired plate which is covered by a likewise unpaired epidermal scute. 



A similar reduction seems to have taken place at the root of the neck. The 

 first of the original 14 marginals is, in Dermatemys and in Ginosternum, still in a 

 debateable condition. It may be a pair of true marginals, or it may represent the 

 pair of costals of the 8th vertebra provided the corresponding neural element has 

 fallen out. At least it seems to explain how by an analogous process the median 

 element of the 9th vertebra, the so-called nuchal, has been produced. Unfortunately, 

 nothing is known as to what might be used as a criterion for determining the nature 

 of these so-called first marginals. The study of the ontogenetic history of Dermatemys, 

 or of Ginosternum, will probably solve this question. 



A later phylogenetic stage would be characterised by the suppression of the supra- 

 marginals, and by the reduction from eight to seven to six and ultimately to even 



The frequency of the Abnormalities. 



