The mandibular foramen lies a little below the level of the alveolar 
border of the body, about 4.5 mm above the ventral border of the 
ramus and about 13 mm behind the alveolus for M,. The foramen 
faces posteriad and somewhat laterodorsad. 
P, and P, are double-rooted, with the posterior root being larger 
than the anterior one. The base of the crown bears cingula 
anterobuccally and posteriorly on both teeth. The cingula are stronger 
on P,; much of the posterior cingulum of P, has been broken away. 
The posterior cingulum of P, is much better developed than the 
anterior one, which was also true for P, judging from the preserved 
base of its crown. The posterior cingular region of the base of the P, 
crown is little deflected linguad. 
The crown base of both P, and P, bears three projections arranged 
one behind the other anteroposteriorly to form a blade compressed 
buccolingually. The blade slightly curves linguad on the most ante- 
rior of the projections, the anterior accessory cusp, in both the 
premolars and on the most posterior projection, the posterior acces- 
sory cusp, in P,. The middle cusp, the protoconid, culminates 
slightly anterior to the half of the tooth length and is distinctly 
largest, whereas the anterior accessory cusp is smallest. The anterior 
and posterior accessory cusps are stronger and larger relative to the 
protoconid on P, than they are on P,. The cusps are divided by 
prominent v-shaped notches on P.. In both teeth, the tips of the 
anterior and posterior accessory cusps are noticeably worn away 
exposing the dentine. Wear has also broken through the enamel at the 
tip of the protoconid on P,,, but over a very small area only. 
The crown of M, is supported by two strong roots. There is no 
cingulum on the talonid, but there are two cingula running along the 
buccal base of the trigonid from the anterior end of the paraconid to 
the most anterior portion of the protoconid buccally and to that of the 
metaconid lingually. The buccal cingulum is very strong, whereas 
the lingual one is poorly developed. 
The trigonid is notably arched buccad, making its lingual contour 
concave when viewed occlusally. The carnassial blade comprises the 
buccal ridge of the paraconid and the anterior ridge of the protoco- 
nid, which are divided by a deep, slit-shaped carnassial notch. The 
carnassial blade is rather deeply worn exposing dentine. The shear- 
ing surface on the buccal side of the paraconid and protoconid is 
considerably worn. Viewed from the occlusal surface, the carnassial 
edge of the paraconid abruptly turns anteriorly into a long, trenchant 
lingual ridge descending towards the metaconid from which it is set 
off by a valley. The carnassial edge of the protoconid curves 
posteriorly at an obtuse angle to continue into a sharp, partly 
damaged ridge that descends obliquely until it meets the metaconid. 
The anterior and lingual ridges of the protoconid delimit the lingual 
wall of this cusp, which flanks posterobuccally a deep, spacious 
valley that sets the protoconid off from the paraconid and metaconid. 
In addition to the anterior and lingual ridges, the protoconid exhibits 
a very short ridge, which is mostly worn away, on the base of its 
posterior wall. This short ridge ascends occlusolinguad from the 
anterior end of the anterior edge of the hypoconid. There is an 
extensive wear facet on the posterior surface of the protoconid. 
The metaconid is stout, well detached from the protoconid, and 
proportionally short anteroposteriorly. In lingual view, it resembles 
an isosceles triangle with its anterior and posterior profiles being 
slightly convex. In posterior view, the lingual contour of the metaco- 
nid is also slightly convex. A small part of the metaconid projects 
posteriorly beyond the protoconid so that its posterior edge is visible 
in buccal view. The slopes of the metaconid are angulated anteriorly, 
buccally, and posteriorly into ridges of which the buccal ridge is 
most trenchant or sharpened and the posterior one is most rounded or 
blunt. The anterior ridge descends towards the lingual ridge of the 
paraconid, from which it is separated by a valley. The buccal ridge is 
M. WOLSAN AND M. MORLO 
united with the lingual ridge of the protoconid at a prominent, V- | 
shaped notch. The posterior ridge meets the lingual wall of the 
talonid. The occlusal part of the posterior surface of the metaconid is 
worn. 
Viewed occlusally, the posterior wall of the trigonid is almost ~ 
straight, while the buccal and lingual contours of the crown are — 
concave at the area where the trigonid meets the talonid. The buccal | 
concavity is much better marked than the lingual one. ! 
The talonid is deeply basined. Its buccal wall consists of an | 
anteroposteriorly elongate hypoconid, which is the largest cusp on 
the talonid. The hypoconid is buccolingually wider and has its outer 
surface more inclined than is the case for the lingual wall of the | 
talonid, making the talonid basin appear to be shifted linguad in | 
occlusal view. Although wear has breached the enamel along the © 
hypoconid, it is evident that the tip of this cusp was originally 
situated within the posterior half of the cusp length. The hypoconid 
is detached from the posterior wall of the talonid by a distinct V- 
shaped notch that is continued into an occlusobasal groove on the | 
outer surface of the talonid. 
The posterior wall of the talonid is lower than the buccal and_ 
lingual walls. It is somewhat worn occlusally and produced into — 
three low, poorly differentiated elevations. ; 
The lingual wall of the talonid forms two projections separated 
from each other by a notch. The anterior of these projections, the © 
entoconulid, is small, whereas the posterior one, the entoconid, is | 
much larger, being the second largest cusp on the talonid. The tips of — 
both the cusps are worn, exposing dentine facets. | 
TYPELOCALITY. Although Pomel (1847) did not indicate the place \ 
of collection of the name-bearing type of *Plesictis’ croizeti explic- 
itly, it is obvious from the contents of his article that the specimen 
had been excavated from Tertiary deposits in the region of Vaumas 
and Saint-Gérand-le-Puy (Allier, France). Several years later, Pomel 
(1853: 97, 1854: 61) expressly attributed that fossil to the Tertiary 
sediments of Langy (‘Terrain tertiaire 4 Langy’), which is a village | 
situated about 3 km west of Saint-Gérand-le-Puy and some 25 km) 
southwest of Vaumas. Most subsequent authors listed the locality as) | 
‘Saint-Gérand-le-Puy’. It deserves to be noted, however, that the 
name Saint-Gérand-le-Puy has generally been applied in the litera-/ 
ture to encompass various fossil sites discovered in several quarries i 
in the region of the village Saint-Gérand-le-Puy, including the! 
locality Langy as well (Cheneval, 1983). 
The only published statements about the type locality of “Plesictis’, 
croizeti that were significantly different were those of Gervais)” 
(1859) and Dawkins (1880a—b). According to the former author, the): ‘ 
holotype of ‘Plesictis’ croizeti was found in calcareous marls of) ~ 
Miocene age in the environs of Issoire in the department of Puy-dey 4 
Dome (his p. 250: ‘Fossile dans les marnes calcaires de étage) 5 
miocéne aux environs d’Issoire (Puy-de-D6me)’). Dawkins in C : 
a 
tte 
statements (1880a: 386, ‘Issoire, Volvic (Puy-de-Déme)’; 18805: 
505, ‘Issoire, Volvic, Puy-de-dome [sic]’) simply quoted Gervai 
(1859). However, neither Gervais nor Dawkins presented any sup ¥ 
porting evidence for their assertions. 
Gervais’s (1859) referral of the holotype of ‘Plesictis’ croizeti to) ~ 
the locality Issoire is rather intriguing since that author was appar 
ently familiar with Pomel’s (1847, 1854) papers as indicated by theil) 
citations in his work, and since he studied that fossil during his visi 
to the British Museum (Natural History) shortly after it had beer 
acquired by that institution, which is evident from footnote 2 on p| ) 
11 in Gervais (18525). In addition, one of the two labels on the typ¢ 
specimen of ‘Plesictis’ croizeti, which lies in its box and refers ittd 
‘Herpestes croizeti’ (the only other label on the fossil is its registe 
number), identifies the holotype as coming from the ‘Uppel )), 
~ 
iL 
r 
