1870.] . OWEN—CHINESE FOSSIL MAMMALS. 419 
one may make out in the reduced figure (half natural size) the 
presence, but not the number, of marginal mamille on the fourth 
or penultimate ridge. But whilst this figure illustrates the degree 
of correspondence, it serves at the same time to show the specific 
degree of difference between the second upper milk-molar of Mas- 
todon (Stegodon) sinensis and Mastodon elephantordes, Clt., or Stegodon 
insignis, Fr. 
The original of Falconer’s figure being in the British Museum, 
yields the following admeasurements, giving the difference in size 
and proportions between :— 
Steg. sinensis. . Steg. insignis. 
in, lines. in. lines, 
Length of crown... ... ... ... 2 10 2 6 
Greatest breadth of crown... ... 2 0 2 1k 
In Stegodon sinensis the siete tooth is longer (antero-posteriorly ) 
in proportion to its breadth. The tubercle at the outer interspace 
between the first and second ridges is larger and situated more 
immediately upon the interspace, closing it externally. The trans- 
verse divisions increase in breadth from the first to the fourth, the 
last being narrower, though not to the same degree as in Steg. in- 
signis; neither do the ridges increase so regularly to the fourth in 
the Ava and Siwalik species as in Steg. sinensis. The ridges in Steg. 
insignis are loftier as well as narrower than in Steg. sinensis; the 
sculpturing is somewhat coarser; the fourth unworn ridge shows 
about twelve mamille. The fangs or roots of the tooth are not 
defined in the Siwalik tooth. In the Chinese one the fore part of 
the crown, divided externally into the two anterior ridges (Pl. I. fig. 
2,1 & 2), is supported by one fang deeply grooved on the side 
turned toward the second: this root is much larger, and supports all 
the remainder of the tooth; its base being entire, we are able to 
infer that only the hind part of the last coronal ridge (5) has been 
broken away, and that there could not be any added talon at that 
end of the tooth. The anterior vertical surface of the crown (ib. 
fig. A) shows a smooth concavity with the enamel worn through to 
the dentine by the pressure against the antecedent molar (d 2). The 
thick layer of cement which coats the dentine of the base of the 
tooth and its two divisions is well preserved, which leads me to 
doubt whether it could at any time have existed in the coronal clefts 
of the teeth of the present species. 
The molars of Elephas (Stegodon) insignis, Fr., chiefly differ from 
those of HE. (St.) Chiftri, Fr., in the much greater mass of laminate 
cement that fills up the valleys*. As this difference is to be added 
to those above described, it further opposes the reference of the 
Chinese proboscidian tooth to the St. imsignis. 
Of the Elephas (Stegodon) Oliftti, Fr., no tooth has yet been 
acquired homologous with the Chinese specimen. From the general 
analogy of retention of pattern of grinding-surface, notwithstand- 
ing the increase of size and number of coronal ridges as the molars 
recede in position, we may certainly infer, from there being no mark 
* Falconer, ‘ Paleontological Memoirs,’ 8vo, vol. ii. p. 85. 
