282 DR. H. W. MARETT TIMS ON 
dificult to imagine how shortening of the jaws could have had 
any effect in bringing about a fusion of teeth of different denti- 
tions; nor, indeed, can one perceive any other change which would 
produce such an effect. The suggestion I would offer is, that the 
three longitudinal rows of cusps are due to the primitive cones 
with cingulum-cusps developed to their inner and outer sides 
respectively. It may be objected, that these inner and outer 
secondary cusps are as pronounced and of equal size as the 
central primitive cones in the true Multituberculata. I do not, 
however, consider this to be any great difficulty, since the Multi- 
tuberculata must have been extremely specialized animals, as is 
shown by their dental formule ; and, moreover, a very similar 
condition of the cusps is to be seen in the molars of existing 
frugivorous bears. Hach molar tooth of the Plagiaulicide and 
Polymastodontide, in which there are three longitudinal rows of 
cusps arranged in numerous transverse rows, would consequently 
represent an antero-posterior fusion of several teeth with their 
external and internal cusps. In other members of these families, 
for example Bolodon, in which the molars bear only two antero- 
posterior rows of tubercles, one of the three rows is non-developed. 
From a comparison with the teeth of existing mammals, I am 
inclined to believe that the series in this form which is wanting 
is that of the external cingulum, it being quite exceptional to 
find this series well-developed, though it is to be found in some 
of the Insectivora and in Ofocyon among the Canide. This 
conclusion receives some confirmation from a comparison with 
the teeth of several species of the Polymastodonting. In a paper 
by Osborn and Earle [13] describing these, they state that in 
P. taoensis, “ although the lower molars typically exhibit but two 
rows, we occasionally observe a postero-external accessory row 
upon the first and second molars;” and again, “ the comparison 
with Meniscoessus shows an average addition of two cusps to the 
first molars in both jaws, and an apparent degeneration of the 
outer row in the second upper molar.” 
In the course of the development of the molars in the guinea- 
pig, the three longitudinal rows of tubercles are present as a 
transitory condition, the external cingulum disappearing giving 
rise to a tooth with but two antero-posterior rows of tubercles 
(Pl. 26. fig. 6). 
The next point to which I would refer is, the similarity of 
dentitions found in the Rodentia and Multituberculata. In both, 
