502 Dr. J. Allan Thomson—On the Terebratellide. 
species to which I have access, Dallina floridana,' possesses hinge- 
plates excavate anteriorly, under which the septum reaches right to 
the umbo as in Magellania, but there is this difference, that in Dallina 
jfloridana the crural bases are not closely applied to the inner sides of 
the socket ridges, as is the case in Magellania flavescens, but run 
independently from the umbo, so that each hinge-plate is separable 
into two parts, one between the socket ridge and the crural base, the 
other between the crural base and the middle line over the septum. 
It is this distinction, so clearly marked in this species, that led me 
to revive the term of ‘‘ socket ridge’ for the ‘‘ buttress’’ forming the 
inner wall of the dental sockets. In most types of cardinalia the 
crural bases are firmly united with the inner sides of the socket ridges 
and cannot be separately distinguished, although sometimes, as in 
Neothyris lenticularis, even when the two are firmly united laterally 
a line of demarcation can be more or less traced on the upper surface. 
Should Dallina septigera possess the same features as D. floridana, the 
type of cardinalia above described may be termed the Dalliniform 
type. In looking for a forerunner of Dallina with Terebratelliform 
loop, one would require as an essential a Dalliniform type of eardinalia. 
This may exist in TZerebratula spitzbergenensis, Davidson, which 
Jackson groups with Dallina septigera and D. floridana in type of 
cardinalia,:-but these species appear to have distinct beak characters. 
In the terminology of Buckman (1916) Dallina had a mesothyrid 
foramen,while Zerebratula spitzbergenensis appears to have, according 
to Davidson's figures, a submesothyrid foramen. More probably 
the forerunner of Dailina looked for will be found in Terebratella 
Marie, A. Adams, or in the Italian Pliocene form Z. septata, 
Philippi. 
In my previous paper (1916, No. 1) I suggested that the forerunner 
of Macandrevia with Terebratelliform loop might be looked for in 
Terebratula frontalis, Middendorff, but refrained from creating a genus 
for its reception because of ignorance of its hinge-characters. Jackson 
states that it possesses somewhat obscure dental plates much as in 
Terebratalia, and that its loop development and cardinalia suggest 
relationship with the Terebrataliform stage of Macandrevia cranium. 
I have now secured a specimen of this species and find some further 
points of agreement with Macandrevia, but also some differences. 
Jackson, in describing the pedicle collar, mentions that it is never 
developed in the higher long-looped forms, in which Macandrevia is 
included, but stated that in some of these ‘‘there is occasionally 
a thickening in the umbo around the foramen assimilating a pedicle 
collar, but it is fused to the shell and never free anteriorly”’. In 
Macandrevia such a plate is well developed in old shells, and extends 
as far forwards as the dental plates, with the base of which it is 
firmly fused, giving the appearance that the dental plates and the 
pedicle collar form a single structure free laterally from the walls of 
the shell but closely applied to the floor. I have observed this 
feature most clearly in a new species of Jlacandrevia from the 
' Two specimens presented to the Dominion Museum, Wellington, by the 
United States Deep-sea Dredging Expedition off the Coast of Mexico, 1869. 
