IMPROVEMENT IN TROTTING HORSES 207 



whether tending to increase or decrease, and so on. The alleged 

 inheritance of the direct effects of mutilations, injuries, and the 

 like is discussed separately in §§ 10 and 11. 



Improvement in Trotting Horses. — Over a hundred years 

 ago (1796) the utmost speed of the English trotter was stated 

 at a mile in 2 min. 37 sec. Since 1818, accurate records have 

 been kept, which show a gradual increase decade after decade 

 in the speed and in the percentage of swift trotters. The 

 standard has risen and the breed has improved. The mile can 

 now be run in 2 min. 10 sec, or less. It is claimed by Cope 

 and others that we have here direct evidence of the trans- 

 mission of the structural results of exercise. 



Brewer (cit. Cope, 1896, pp. 426-30) relates that about 1818 

 the record speed of the trotting horse was 3 min. to the mile ; 

 in 1824 it was reduced to 2 min. 34 sec. ; in 1848, to 2 min. 

 30 sec. ; in 1868, to 2 min. 20 sec. ; in 1878, to 2 min. 16 sec. ; 

 in 1888, to 2 min. n| sec. ; and finally to 2 min. 10 sec. 

 " The gain in speed has been cumulative. ... It has gone 

 on along with systematic exercise of special function in suc- 

 cessive generations ; . . . there is nothing that would lead us 

 to even suspect that the changes due to exercise of function 

 had not been a factor in the evolution ; . . . there is every 

 appearance and indication that the changes acquired by in- 

 dividuals through the exercise of function have been to some 

 degree transmitted, and have been cumulative, and that this 

 has been one factor in the evolution of speed." 



It is impossible to prove the negative above suggested — 

 namely, that function has not been a factor ; but the affirmative 

 position is robbed of all cogency by the admitted occurrence of 

 rigorous artificial selection. The improvement supposed to be 

 entailed ma} 7 not have been a modification at all ; but, supposing 

 it was, the interpretation of the result simply by the hypothesis of 

 use-inheritance gives a false simplicity to the case. It overlooks 

 the selective breeding which increases the constitutional swiftness, 



