105 
(fig. 13, diagr. 14) the heel of p 4 is likewise bicuspid, and this is 
one of the reasons for the assertion of Lydekker and Lan- 
kaster ( u On the Affinities of Æluropus melan.'\ Trans. Linn. Soc. 
1901) that it is more nearly related to 
Ælurus than to Ursus, From a cur- 
sory glance at p 4 of Æluropus it 
seems also very much like that o 
Ælurus , but a more careful examina- 
tion shows an important difference. 
I am very much indebted to Mr. Old- 
field Thomas for allowing me during 
a visit to the British Museum to detach 
p 4 of Ælurus and Æluropus in order to compare them. In the 
normal condition cusp 6 is supported by a special root which is 
expanded to support the anterior part of cusp 7 as well, when 
this is present, whereas 
its posterior part rests 
on a part of the root 
of 3 or 5; this condi¬ 
tion is seen in all bu- 
nodont raammals. The 
same is the case with 
Ælurus (Sg. 12), and 
here the root under 6 
and 7 is very strong; 
but Æluropus (fig. 13) 
shows the strange con¬ 
dition of the heel-root 
supporting, not the an¬ 
terior heel-cusp which 
is supported by the an¬ 
terior outer root (r 4 --), P 4 s ^ u an( ^ isolated, from the inner and 
ri,, . . ° „ the front side. Note the weak root supporting 
bat the main part of cusp 6 (r§)> and tjhe ^ 3trong . outer roots 
the posterior heel-cusp, (r 4 _ 6 and r 8 ). bi. 
Fig. 12. Ælurus fulgens. 
p 4 isolated, from the inner and 
the front side. Note the strong 
root supporting cusps 6 and 7. 
bi. 
