Teeth in the Centetide. 539 
contrary to what is stated by Leche *, younger specimens 
of Z,.1 and J,2 always possess a well-developed posterior 
basal cusp. 
In the case of Oryzoryctes, Microgale, and Limnogale P.1 
may at once be distinguished from P.d. 1 by the fact that it 
resembles J/.1 much less than the latter does, which is 
especially evident in the condition of the external cusps; the 
foremost of these lies, in the case of P.d.1 and the two 
anterior molars, in the same longitudinal row as those 
following it; in the premolar it is set higher up and is 
separated by a greater gap from those behind. As already 
mentioned, this difference is much less distinct in the case of 
Centetes, Ericulus, and Echinops. 
The number of the external cusps on fresh cheek-teeth of 
the molar series is from four to five in AMicrogale, on P.d.1 
usually five ; in Oryzoryctes five on P.d.1, as on M.1 and 
M.2. These external cusps are divided by a median gap 
into an anterior and a posterior series ; where five cusps are 
present, there are in some cases two cusps in the anterior and 
three in the posterior division—in others the reverse. In 
P.1 only three external cusps are found; in the case of 
Microgale the median one, in comparison with that on either 
side of it, has become disproportionately large and _ lofty. 
With reference to the various species of Microgale the follow- 
ing remarks must be made. J/. Dobsoni: P.1 corresponds 
in its carnassial-like condition pretty closely with the P. 2 of 
Ericulus, and similarly P.2 of M. Dobsont is almost iden- 
tical with the P.3 of Hriculus; at the most, in the case of 
the former, the internal cusp is somewhat more strongly 
developed; nevertheless in the latter also the tooth, as 
already mentioned, has but two roots. P.3 is a secodont 
tooth with two roots ; this applies to all species of Mfcrogale. 
—In WM. Cowant P.2 differs but little from P.1; in height 
especially the two teeth are about equal, while in MZ Dobson 
P.2 is much lower than P.1. P.2 is only slightly more 
compressed, and in consequence somewhat more elongate ; 
the ‘protocone”’ is developed to precisely the same extent 
as on P.1. A much greater amount of compression is 
shown by P.2 in the case of M. Thomasi, in which also the 
‘6 nrotocone ”? is unusually reduced in comparison with that 
of P.1. In the genus Microgale the maximum of com- 
pression (parallel to the longitudinal axis) is found in the 
case of M. gracilis; nevertheless a feebly developed “ proto- 
cone ”’ is still present on P.2, while in the case of the tooth 
* Loc. cit. p. 519. 
36* 
