90 
anything the tritubercular tooth must be supposed to have originated 
by concrescence; but. the further development has taken place through 
differentiation and -cusp-addition. Certain anomalous circumstances 
as regards the human teeth tend to show that each of the primary 
cusps may contain several reptilian teeth. 
As far back as 1880 Huxley ("Cranial and dental characters 
of Canidæ”) suggested that the evolution of the complicated molar 
forms is' shown by the successive complication of premolars and 
molars of Centetes, from which type the teeth of Canidæ may be 
derived. Upon different grounds Schlosser supported a similar 
theory ("Die Entwickelung der verschiedenen Såugetierzahnformen im 
Laufe der geol. Perioden”, 1891), viz., that in Creodonta, 'Condyl- 
arthra and Ungulata the outer side of both premolars and molars 
is the more primitive, and the inner side is secondarily complicated. 
This "Premolar Analogy Theory” is also advocated by Wortman 
(1902—03) who emphatically asserts that the cusp-evolution in 
molars is exactly the same as that seen in the premolars, where it can 
actually be followed step by step from the haplodont into the sexi- 
tubercular stage; on this theory the paradox is also withdrawn that 
the reptilian cones are on the outer side of the premolars and on 
the inner side of the molars in the upper jaw, and the embryolo- 
gical and palæontological testimonies as to the primary cone are in 
harmony with one another. 
In "Evidence bearing on Tooth-Cusp-Development” 1906 Gidley 
contests the Cope-Osborn theory and maintains a polyphyletie evo- 
lution of the mammalian molars. Osborn has supported his theory 
chiefly upon the teeth of Triconodon and Dryolestes; but he has 
overlooked a very important circumstance. The teeth of Triconodon 
consist of three nearly equal cusps with an outer and an inner 
cingulum and supported by two roots. Dryolestes has in addition 
to the three cusps noted by Osborn (pr., me., på.), two others 
one external cusp (pa) and the other a small intermediate cusp 
between pr. and me.; there is thus a break between 'the two genera 
and the transition is not made easier by the teeth of Triconodon 
