ON THE PHYLOGENY OF THE CARAPACE, ETC. 801 
7. In the primitive upper Triassic tortoise Proganochelys there was 
present a series of supramarginal bony plates (fig. 6), discontinuous in 
the middle but well developed anteriorly and posteriorly.** In the 
very interesting Upper Triassic form Stegochelys, Jaekel (1914) found 
a continuous series of eleven supramarginals on each side of the 
shell. If we compare these forms with the Leatherback and Archelon, 
: becomes probable that these elements also belong to the epithecal 
ayer. 
8. It is well known that the horny scales of reptiles often corre- 
spond with bony scutes. Hay and Newman, believing that horny 
scales develop in reptiles over bony scutes only,** concluded that the 
well-known horny shields of the Testudinata must also originally have 
developed upon the surface of bony scutes. As the shields are arranged 
differently from the bony plates of the thecophorous shell, they must 
have developed upon other ossifications, which can only have been 
“epithecal.’ Of course, we need not assume that these epithecal bony 
plates were as large as the horny shields are now, as the latter may 
have enlarged far beyond the border of the corresponding bony scutes. 
It is true that the bony marginals in Testudinata, though here supposed 
to belong to the epithecal layer, do not correspond to the marginal 
horny shields, but break joints with them.** Yet there seems to be 
some connection—e.g., in case of abnormalities of the marginals the 
marginal horny shields always are abnormal too, and the reverse is 
also always true, whilst no such relation is found between the thecal 
elements and the overlying shields.*® It is probable that in the very 
primitive Proganochelys and Stegochelys the bony marginals and 
supramarginals corresponded exactly with the overlying horny shields.*” 
We cannot directly control the correspondence of bony and horny scutes 
in the Leatherback, as it has lost its shields in the adult and has not 
yet formed the osseous mosaic shell in the new-born specimens, whilst 
the intermediate ages are unknown; however, the young Dermochelys, 
as it leaves the egg, is covered with very small numerous horny shields, 
showing the same arrangement, with the typical rows of larger elements 
on the keels, as is found in the bony epithecal scutes of the adult. 
The supramarginal horny shields of Macroclemmys, and of the 
extinct Boremys (Amphichelydia) are in this light also an indication 
of the former presence of bony supramarginals; and the inframarginal 
horny shields that are so widely distributed in the Testudinata point 
to the former development of an inframarginal series of bony elements. 
The interpretation of Volker and Menger is in this important matter identical with 
the older view of Goette. But it differs essentially in that the costal and neural 
plates are considered parts of the inner skeleton by both Goette and Jaekel, whilst 
Menger has found what seems to me conclusive evidence, that they also are dermal 
ossifications, belonging to the same deeper layer as the nuchal and suprapygals. 
Compare Gegenbaur, 1898, p. 175-176. 
83 Jaekel, 1906, p. 66, who calls these elements submarginalia. 
84 This does not seem to be a hard and fast rule, as there is important evidence 
of the formation of horny scales independent of dermal ossifications in Lacertilia 
(Schmidt, 1910, p. 640; 1912, p. 86). 
85 This difficulty did not present itself to Hay and Newman, as they considered 
the marginals as belonging to the thecal layer. 
86 Newman, 1906, p. 96. 87 Jaekel, 1906, p. 66; 1914. 
1913. 3 F 
