622 The American Naturalist. ` a 
of the pig as a modern bunodont, and as selenodont types a number 
of elk (Alces palmatus) embryos, one stage of the deer (Capreolus 
caprea), abundant material of the ox and sheep as hypselenodont 
types, and of even greater interest, an embryo of one of the rare and 
transitional ancient group of Tragulide. As, in course of his investi- 
gation, he found abundant evidence in support of the remote trituber- 
cular origin of these modern highly modified teeth, he adopts Cope’s 
theory and Osborn’s nomenclature and homologies, so that the reader 
can readily compare his plates and descriptions with the palæonto- 
logical series as recently figured by Cope, Scott, Lydekker, Schlosser 
and others. 
We may now quote his conclusions: “Asa result of my investiga- 
tions it is evident that both the bunodont Suide, and selenodont rumin- 
ants present a closely similar initial bunodont stage; according to 
which the Cope-Osborn opinions are in general conformed by embryo- 
logical data, although in many points, especially as regards the (more 
elevated position of the) protocone, ontogenesis is no longer parallel 
with phylogenesis. 
2. The differentiation soon follows whegeby the separate cones and 
conids in the pig transform into pyramids and in the ruminants into 
crescents. 
3. The transformation of the cones affects the separate cones in suc- 
cession, not beginning with the protocone but with the earlier devel- 
oped paracone. [The author has unconsciously strengthened his proof 
here, for it is well known that in the fossil series, the external upper 
cusps, i. e., the paracone and metacone, assume the selenoid form earlier 
and more universally than the protocone. ] 
4. In the last two upper deciduous premolars (D* D’) from an 
originally conic metacone is developed a tooth with two cones, by the 
addition of the paracone; then, in D* at least, appears the protocone 
and finally the hypocone. In D’ the protocone is the last cusp added. 
[Here again the author demonstrates an approximate parallelism be- 
tween the ontogenesis and phylogenesis; he is evidently not aware of 
the observation of Schlosser that the upper premolar cusps do not 
appear in the same order and are therefore not homologous with the 
molar cusps. The actual order of evolution of the premolar cusps 
was observed by Osborn’ in Hyracotherium as follows: A cusp 
analogous with the paracone (really the protocone) first appears, 
then the outer cusp divides into two, paracone and metacone, 
1 MSS. of chapter upon the Æguide for the “ American Fossil Mammals.” 
asap ge eat ee ae OTE Se ATN E TERE eae et a: a See ere ge Teeter Mee Pee ee ee 
Se eee ee ge 
DS ae ye Re EE ne Rede Ce re ee NA 
RO ig ea T e eo 
