24. CONTRIBUTIONS TO PALASONTOLOGY 
inized” form of the upper canines in man, together with the lack of an ob- 
lique anteroexternal face on his anterior lower premolar, to be entirely 
secondary features, early acquired (and showing already even in Sinan- 
thropus) but nevertheless secondary. 
The crown of the anterior lower premolar varies from a laterally com- 
pressed, almost secant form (associated with the shearlike blade of the 
upper canine) to an obliquely placed, more oval, almost bicuspid form. 
Similar differences may be seen in modern anthropoids. 
The posterior lower premolar varies but little, chiefly in the degree of 
development of the talonid. 
The upper premolars, both anterior and posterior, have become com- 
pletely bicuspid, as in modern apes. The anterior usually differs from the 
posterior premolar in that the inner (lingual) cusp is relatively much lower 
and the outer higher, while in the posterior premolar the two main cusps 
are almost equal, the anterior premolar (P3 of the primitive placental 
dentition) having a high pointed labial cusp which cooperates with the 
posterior (distal) blade of the canine to produce a sharp jag or notch 
analogous to that seen in the carnassial teeth of the dentition of carnivores. 
Ramapithecus, however, is an exception to this rule, since its anterior upper 
premolar is in a relatively advanced stage. The inner lobe of both the 
upper premolars is bordered by two parallel transverse ridges. 
The lower molars vary widely, from the long narrow crowns of Sugriva- 
pithecus gregoryi (pl. 8E, breadth index of M2, 76.7) to the very broad 
crowns of Bramapithecus thorper (pl. 8D, index, 106). The several elements 
of the molar crown vary widely in position, height, etc., thus producing cor- 
responding differences in the details of the general “Dryomthecus pattern.” 
Formerly minor surface features of the lower molar crowns have frequently 
been used as specific characters. Our review of the material shows, how- 
ever, that these minor features are often merely due to the degree of wear. 
The most variable tooth, both in proportions and in dimensions, is the 
third lower molar, which is relatively very large in the type of Sivapithecus 
himalayensis and very small in the type of Bramapithecus ? sivalensis Lewis, 
very short and relatively wide in Bramamthecus thorpei, exceedingly nar- 
row in Sugrivapithecus ef. gregoryr. If we may judge by the excessively wide 
variability of M3 in recent anthropoids and man, the variations of this 
tooth in Siwalik anthropoids may eventually prove to be of minor system- 
atic value. But in this question as in so many others, the great need is for 
“more fossils and always more fossils.” 
With regard to the general construction of the jaws and teeth the Siwalik 
anthropoids, taken as a whole, were on a distinctly infrahuman grade of 
evolution. But in particular features, now one and now another approaches 
or even overlaps the outer zone of human variations. For example, the 
lower second and third molars of Bramapithecus thorpei Lewis in respect to 
the breadth indices even surpass some human teeth; and in earlier papers it 
