xxx. iz RATE OF EVOLUTION OF HORSES 739 



assessed as a comparable entity throughout (a large assumption, this!) 

 and dividing the eight genera (excluding Eqiiiis) on the direct line into 

 the time involved between *Hyracotherium and *Pliohippns (50 

 million years), we should have 6-3 million years per genus. However, 

 there is reason to suppose that individual genera lasted for very 

 different times, *Miohippus, for instance, less than a third of the time 

 of *Merychippus. Therefore if the criterion of a genus is constant, the 

 rate of evolution must vary. 



Sufficient fossil horse material is available to allow consideration 

 whether known rates of mutation are likely to be adequate to account 

 for the observed evolutionary changes. Simpson calculates that from 

 * Hyracotherium to Equus there must have been at least 15 million 

 generations, which, with a population in North America of 100,000 

 (a low estimate), gives a total of 1-5 x io 12 individuals in the 'real and 

 potential ancestry of the modern horse'. One in a million is a moderate 

 rate for large mutations at any locus in Drosophila, and this would 

 give 1 -5 million such mutations for a single locus in the horse ancestry. 

 It would be safe to assume that one-fifth of these (300,000) were in the 

 direction favoured by selection and that one-tenth of all such genes 

 affect a structural change, such as ectoloph length. The actual increase 

 in this length between * Hyracotherium and Equus was from 8 to 40 mm 

 which, divided into 300 steps, gives an increase per mutation of 

 only o-i mm. This is a reasonable figure, and such calculations 

 suggest that observed mutation rates are quite adequate to account 

 for the evolutionary changes, even neglecting possible multiple actions 

 and interactions of genes, by which the speed of evolution could be 

 further increased. 



12. Conclusions from the study of the evolution of horses 



Careful consideration of the fossil horse material therefore shows 

 reason to suppose that evolution has proceeded by gradual change. 

 As more and more evidence becomes available the series becomes 

 more and more complete, and incidentally the nomenclature increas- 

 ingly confusing. Incompleteness of material may give an impression 

 of evolution by jumps and saltation, especially when, as in the Old 

 World horses, there have been successive migrations into one region 

 from another. The European palaeontologists, finding *Hyraco- 

 therium, *Anchitherium, *Hipparion, and Equus, without intermediate 

 forms, interpreted the evidence as showing evolution by saltation. 

 This was indeed a reasonable deduction from the facts, but was not 

 the only possible one, as has since been shown by the discovery of 



