103 
affirming its nature as tke true 4 (fig. 4 and 5). In Canis (fig. 6 
and diagr. 7) it also occupies tke last mentioned place, but is muck 
smaller tkan 1 and distinctly separated from it, wkereas it nearly 
has coalesced witk 5. This coudition is still more advanced in 
Fig. 5. Hyæna crocuta. 
dp 3 (from above and from tke 
outer side). 1 /i. 
Fig. 6. Canis vulpes. 
dp 3 (from above and from tke 
inner side). 5 / 2 and 1 ji. 
Bassaris (fig. 7) wkere 4 has nearlj lost its individuality, thougk 
1 is still large. — 6 being connected witk 4, not witk 1, it is 
evident tkat 1 being large and 4 small the anterior part of tke tooth 
must reack some distance in front of 6, so tkat tke keel is found 
in its characteristic place.—Ckaracteristic of dp 3 is tke relatively 
large 1 and tke rudimentary, sometimes vanishing 4. 
i ¥ 
Fig. 7. Bassaris astuta. Fig. 8. Genetta tigrina. 
dp 3 (from above and from tke p 4 (from above and from tke 
outer side). 3 /i and 1 /i. outer side). 3 / 2 and 1 /i. 
p 4 of a Viverroid e. g. Genetta (fig. 8, diagr. 8), differs from 
dp 3 in cusp 1 being as it were pusked into the anterior outer 
corner, whereas 4 has increased; 5, 6 and 3 are also relatively 
stronger than in dp 4 . In Felis (fig. 9, diagr. 9) cusp 4 has grown 
into a very important cusp, sometimes witk a rudimentary cingulum- 
like shelf on its anterior border, evidently tke remnants of 1. In 
Machærodus (fig. 10) 1 is a little more distinet, but very insignificant. 
In Hyæna (especially H. striata fig. 11, diagr. 10) 4 nearly 
