752 



SCIENCE 



[N. S. Vol. XXVII. No. 697 



in a broad-skulled or long-skulled direction, 

 respectively. 



The important point to note is that the de- 

 scendants of a single ancestral titanothere or 

 of any other vertebrate may become either 

 brachycephalic, mesaticephalic or dolichoceph- 

 alic; in other words, the primitive mesaticeph- 

 alic ancestral form of slcull does not control 

 the form of shull which may he derived from 

 it, yet an evolution tendency once estahlished 

 is pursued to its limits. 



2. Tooth Evolution through Rectigradations 

 or Orthogenesis 



Here, in contrast to the foregoing cases of 

 brachycephaly and dolichocephaly a law of 

 hereditary ancestral control appears to be in 

 operation. The diagrams in Fig. 3, A, B, 

 represent the origin of cuspules in two inde- 

 pendent families of Perissodactyla which also 

 have sprung from a very remote common 

 ancestor. 



The types selected are (4) Orohippus, a 

 Middle Eocene horse, and (B) Palceosyops, a 

 Middle Eocene titanothere. The teeth repre- 

 sent the seven grinders of the lower jaw 

 viewed from the internal aspect. Circles sur- 

 round the new cuspules, which are appearing 

 on the inner sides of these teeth. We observe 

 that cusp for cusp exactly the same cuspules 

 are arising in the jaw of Orohippus as in the 

 jaw of Palceosyops, but that, although an ani- 

 mal of the same geological age, Orohippus is 

 acquiring its new cusps a little more rapidly 

 than Palceosyops, as shown in the following 

 table : 



Palseosyops Orohippus 



', No. of No. of 



Cuapuiea Cuspules 



First premolar 2 2 



Second premolar 2 3 



Third premolar 3 4 



Fourth premolar 4 4 



First molar 3 3 



Second molar 1 1 



Third molar 1 1 



Total .16 18 



This comparison proves that while there is 

 apparently a law of ancestral or hereditary 

 control operating in the genesis of these new 



cusps, and that while the new cusps are ortho- 

 genetic and hence may be termed " rectigrada- 

 tions " (because developing in fixed lines) such 

 law of ancestral control does not determine 

 the rate of evolution of the cu^ps in these two 

 types. The rate of evolution is more rapid 

 in Orohippus than in Palceosyops. 



This observation appears to bar the hypoth- 

 esis that the appearance of these cusps is due 

 to an internal perfecting tendency which oper- 

 ates independently of external conditions and 

 to favor the hypothesis that in some unknown 

 manner external conditions control the rate 

 of evolution, again illustrating the law of the 

 four inseparable factors. 



Conclusion 



The contrast between the origin of changes 

 of proportion illustrated in brachycephaly and 

 dolichocephaly and the origin of new cuspules 

 is, apparently, that the former is independent 

 of hereditary control and not predetermined, 

 while the latter is predetermined or under 

 hereditary control. Both phenomena are con- 

 trolled alike as to rate of evolution by adapta- 

 tion to external conditions, namely, by the 

 kind of food on which the animal subsists. 



These considerations appear to me to sus- 

 tain my hypothesis of the independent opera- 

 tion of two primary factors at least to pro- 

 duce an harmonious adaptive result. 



Henry F. Osborn 



the filling of emerald lake by an alluvial 



FAN 



About four miles northwest of the town of 

 Field, in British Columbia, and separated 

 from it by Mount Burgess, lies the beautiful 

 sheet of water known as Emerald Lake. Situ- 

 ated near the head of a broad glacial valley, 

 this lake has been formed probably by the 

 damming of the original channel by a heap 

 of glacial debris, perhaps supported by a re- 

 sistant outstanding ledge of bed rock at this 

 place. Across this barrier the water has its 

 exit. 



The present interest in Emerald Lake, how- 

 ever, rests not so much upon its mode of for- 

 mation as upon the fact that it is slowly being 



