70 : PERMIAN FOSSILS. 
the imperforate valve ; it is also to be seen both posteriorly and anteriorly attached in 
Magas pumilus ;* while in the singular Pachyrhynchus roseus it is simply united to the 
anterior end of a corresponding plate; whereas in Zerebratula (?) Natalensis it seems 
to be reduced to two erect processes springing from the centre of the imperforate 
valve. 3 
Perhaps the loop, in no case, exists under so enlarged a form as in Stragocephalus 
Burtini,—a shell on which I feel it necessary to make a few observations corrective 
of what I have elsewhere stated respecting its internal structure.” Having lately 
succeeded in exposing the interior of a specimen, which for a long time defied all my 
efforts to disclose, I am now enabled to state, that the mternal structure of this 
species does not consist of two parts folded, so as to form a pair of “‘ symmetrical 
subgyrate appendages,” as I formerly stated ; but it is in reality a single piece, having a 
tolerably close agreement with the loop of certain Verebratulide (Pl. XIX, fig. 1 d). 
The crura have certainly an unusual peculiarity ; their modification from the ordinary 
form, however, is not difficult of explanation. Shorten the ascending stems (c), and 
conjoin them with the descending portions of the loop (¢),—the resulting form will be 
precisely the same kind of projecting crura seen in Pl. VI, fig. 45e, and Pl. XX, 
fie. 12 #. 
The figures which M‘Coy and Davidson have given of the spirals or labial supports 
of Delthyris and Trigonotreta, show very clearly that they are connected with the 
crura, and these again with the crural base near the socket-ridges, as in 7erebratulide ; 
and it affords me much pleasure in finding my own observations on the spirals of 
Delthyris cristata fully corroborating their representations. 
The imperforate or small valve of many genera has a plate holdimg a_ position 
similar to the one already noticed, running along the medio-longitudinal line of the 
opposite valve (Pl. XIX, figs. 34, &c.). To distinguish this plate from the latter, it is 
proposed to term it the dorsal median plate. In Waldheimia Australis (Pl. XX, fig. 11 G) 
and several other species, it supports the crural base; in Camerophoria multiplicata 
(Pl. VIII, figs. 3, 47) it steadies a spatula-shaped process, which will hereafter be 
described; and in MZagas pumilus it affords support, at its anterior end, to a modified 
form of a probably homologous structure ; but in Productus (Pl. XI, fig. 10 4), which has 
neither a crural base nor loop, its use appears only to keep certain muscles separated 
from each other. The dorsal median plate apparently subserves the latter purpose in 
Leptena (Pl. XX, fig. 7 6), and some other shells; but im another form, Strophomena 
Dutertrv, there seems to have been a failure in this respect, since the muscles were 
implanted on lamellar fulcra, which curve over it, and coalesce, forming, as it were, 
a complete arch. 
‘ Vide Note sur le Magas pumilus, par MM. Th. Davidson et Bouchard-Chantereaux, Bulletin de la 
Soc. Geol. de France, 2° série, tom. v, pp. 139-41, pl. u. 
” Annals and Magazine of Natural History, vol. xviii, p. 88. 
