THE ORIGIN OF SPECIES 



B 



p m 



paracone height from arctoid allometry — ^ 



Figure 98. First Upper Molars of Bears; p, paracone, and m, metacone, the promi- 

 nent cusps of the teeth. A and B, small and large Ursus arctos; D and E, small and 

 large U. spelaeus; and C, a tooth from U. arctos with spelioid allometry. The dashed 

 line in E indicates what the height of the paracone would have been had arctoid 

 allometry prevailed. ( From Kurten, Evolution, V. 9, 1955. ) 



are found in U. etruscus, an early Pleistocene bear which was ancestral 

 to both species. Kurten has made the reasonable, but unproved, assump- 

 tion that genes governing the two kinds of allometry are alleles, Aq for the 

 arctoid, more extreme type, and Ag for the spelioid, more moderate type. 

 Both alleles, then, were present in the ancestral U. etruscus. A sample 

 from the mid-Pleistocene, which is probably ancestral to spelaeus and pos- 

 sibly to arctos, shows about 67 per cent A^, and 33 per cent Ag. The Hardy- 

 Weinberg formula leads to an expectation of 4 AgAa : 4 AaA^ : 1 AgA^. In 

 this sample, many molars are intermediate, and hence probably based 

 upon heterozygotes. The actual sample count was 42 arctoid, 50 intermedi- 

 ate, and 8 spelioid, not a significant deviation from the mathematical 

 expectation of 44.45 : 44.45 : 11.1 ( = 100, the sample size). In U. spelaeus, 

 which appeared later, the gene A^ disappeared altogether, while both 

 genes have been retained in U. arctos, but Ag is less abundant, making up 

 only 23 per cent of the gene pool of a recent Finnish population of bears. 

 Thus Kurten has supplied an outline history of a pair of alleles over the 

 past million years. 



Some probable selection forces can be inferred. U. spelaeus was a large 

 bear, and arctoid allometry would have produced a very high tooth, jut- 

 ting out of the tooth row and not working harmoniously with the others. 

 The spelioid pattern in U. arctos, however, leads to a low-crowned tooth 

 which must wear down more quickly than its neighbors. 



Mutation Pressure and Genetic Equilibrium. The above discussion 

 has assumed that no mutations occur, Init this assumption is, of course, 

 not valid. In any particular case, it is possible that mutation might proceed 

 only in one direction, as A— >«, or it might proceed in both directions, 

 A^rt. In the former case, even a slight mutation pressure would eventually 

 lead to a species completely homozygous for the mutant gene, unless a 

 selective disadvantage of the mutant were to prevent this. If the latter 

 were the case, and if all of the genotypes had ecjual selective value, then 

 the frequencies of the two alleles would reach an equilibrium, the nu- 

 merical value of which would depend upon the actual magnitude of the 

 mutation rates in the two directions. 



The equilibrium point is related to the two mutation rates in a simple 



256 



