230 
THE DESCENT OF MAN. 
Paet I.. 
called races. Nevertheless such early races would per- 
haps have been ranked by some naturalists as distinct 
species, so arbitrary is the term, if their differences, 
although extremely slight, had been more constant than 
at present, and had not graduated into each other. 
It is, however, possible, though far from probable, 
that the early progenitors of man might at first have 
diverged much in character, until they became more 
unlike each other than are any existing races ; but that 
subsequently, as suggested by Yogt , 21 they converged 
in character. When man selects for the same object 
the offspring of two distinct species, he sometimes 
induces, as far as general appearance is concerned, 
a considerable amount of convergence. This is the 
case, as shewn by Yon Nathusius , 22 with the improved 
breeds of pigs, which are descended from two distinct 
species ; and in a less well-marked manner with the 
improved breeds of cattle. A great anatomist, Gratiolet,. 
maintains that the anthropomorphous apes do not form 
a natural sub-group ; but that the orang is a highly 
developed gibbon or semnopithecus ; the chimpanzee 
a highly developed macacus ; and the gorilla a highly 
developed mandrill. If this conclusion, which rests 
almost exclusively on brain-characters, be admitted, 
we should have a case of convergence at least in 
external characters, for the anthropomorphous apes are 
certainly more like each other in many points than 
they are to other apes. All analogical resemblances, 
as of a whale to a fish, may indeed be said to be 
cases of convergence; but this term has never been 
applied to superficial and adaptive resemblances. It 
21 4 Lectures on Man,’ Eng, translat. 1864, p. 468. 
22 ‘Die Racen des Schweines,’ 1860, s. 46. ‘Yorstudien fur Ge- 
schichte, &c., Schweineschadel,’ 1864, s. 104. With respect to cattle,. 
see M. de Quatrefages, *■ Unite de TEspcce Humaine,’ 1861, p. 119. 
