1892.] NATURAL SCIENCES OF PHILADELPHIA. 443 



given to the molar cusps would necessarily be transposed. Space 

 fo.rbids, however, a consideration of this question here. In one 

 important respect Rose's views are altogether irreconcilable with the 

 facts of paliBontology, viz., in supposing that complex mammalian 

 molars are formed by the coalescence of simple teeth. The series 

 of genera which appear in successive horizons, and Avhich we may 

 confidently assume to stand in ancestral rehttions to one another, 

 display the entire history of the complex premolars and show that 

 new elements are formed by outgrowth from the old and not by 

 coalescence. Such a series as that of the horses, for example, is 

 sufficient to demonstrate this, unless we are prepared to assume that 

 the apparent connection of the different members of the series is 

 entirely illusory. 



Summary. 



1. Assuming the correctness of Osborn's results as to the 

 homologies of the molar cusps, those of the premolars are differ- 

 ently arranged. In the upper premolars the protocone forms the 

 antero-external cusp. 



2. Additions to the protocone are typically made in the following 

 order : (a) the antero-internal cusp — deuterocone ; (b) postero- 

 external cusp — tritocone ; (c) postero-internal cusp — tetartocone. 



3. In the lower premolars the protoconid takes the same position 

 as in the lower molars, as does also the paraconid, but the meta- 

 conid forms the external cusp of the talon. The places of the 

 meta- and hypoconids of the molars are taken by two corresponding 

 but not homologous elements, the deutero- and tetartoconids respec- 

 tively. 



4. While there is no reason to doubt the homologies of the pre- 

 molar cusps throughout the series, the order in which they make 

 their appearance is by no means invariable, especially in the ante- 

 rior teeth. 



5. In certain selenodont artiodactyls (e. g., Procamelus, Oreodon- 

 tidce) the inner crescent of the anterior upper premolars is not 

 formed entirely by the extension of the deuterocone, but by the 

 coalescence of two ridges, one of which extends from the anterior 

 and the other from the posterior margin of the crown. 



6. The homologies of the cusps in the milk molars agree with 

 those of the premolar elements, but appear to be even more irregu- 

 lar as to the order in which those cusps are developed. Thus, in dj 



