Dr. J. Allan Thomson—On the Terebratellide. 503 
Antarctic, but it also exists in If. cranium.’ It is thus described by 
Dall (1895) in I. americana: ‘‘teeth strong, short, supported each 
by a strong buttress with a recess behind it, and in old specimens 
with a smooth deposit of callus on the surface of the valve between 
the two buttresses.”’ This type of dental plates, supported by a deposit 
of callus on the floor of the valve, differs from that of Hemithyris, and 
may be termed the Macandrevian type. Hemithyris has a true but 
short pedicle collar, with which the deposit of callus in Macandrevia 
does not appear to be homologous. The Macandrevian type of dental 
plates is also found in Zerebratella frontalis. 
. The beak characters of Macandrevia are of an unusual type. The 
pedicle opening consists of two parts—a rounded foramen, which is 
permesothyrid in position,” opéning into an open delthyrium. In the 
usual course of events in Terebratellids, when the foramen has 
attained the mesothyrid position, the delthyrium has become closed 
by deltidial plates. In TZerebratula frontalis the movement of the 
foramen ventralwards does not appear to have gone quite so far, but 
the relation to the delthyrium is the same. 
The cardinalia of Macandrevia are also of an unusual type. The 
crural bases are fused on their outer sides to the socket ridges as in 
Magellania, and from their inner sides two hinge-plates, excavate 
anteriorly, descend obliquely to the floor of the valve, and becoming 
fused with this unite in the middle line of the valve. Inthe Antarctic 
species above-mentioned, from which this description is drawn, these 
hinge-plates do not extend forward beyond the crural processes, and 
there is a raised thread-like line occupying the position of the median 
septum. In Macandrevia cranium, Davidson (1886) states that there 
is no defined cardinal process or median septum, but that two deviating 
septa commence under the umbo and extend toa little more than one- 
fourth the length of the valve. ‘These so-called septa are apparently 
anterior prolongations of hinge-plates similar to those described above. 
It is desirable that the ontogeny of the Macandreviform type of hinge- 
plates should be worked out. It bears some resemblance to that of 
early stages of TZerebratella rubicunda, but it seems not impossible 
that it is a further development from a Magellaniform type. 
In Terebratula frontalis a different type of cardinalia exists. There 
is a small median septum situated in front of the middle of the valve, 
and from it a raised thread-like line extends back to the cardinalia as 
in Macandrevia. The crural bases cannot be traced in the cardinalia 
with certainty. The crura spring from the anterior inner corners of 
strong socket ridges, between which there isa mass of shell substance 
overhanging in front and embayed nearly to the umbo. The outer 
edges of this overhanging mass are raised into ridges, and perhaps 
mark the crural bases, and between these ridges and the socket ridges 
there are well-marked depressions on the mass. On the back of the 
mass is superposed a small transverse cardinal process. 
1 It is figured, but not described, by Fischer & Oehlert (1891, pl. v, 
fig. 10 f.). 
? The beak ridges are poorly defined, and it is difficult to be quite certain of 
their position. The foramen is certainly not of the submesothyrid type, into 
which most Terebratellids with lateral deltidial plates fall. 
