102 SECRETS OF EARTH AND SEA 



if we find that they breed as a rule with the ordinary 

 members of the species. The decisive test is the breeding. 

 If the variety is found not to breed with the regular 

 species, but to keep apart and breed only with other 

 individuals like itself, then we say, "This is no mere 

 variety! It is a distinct species!" Unfortunately we 

 have vast series of animals, insects, and others, from all 

 parts of the world, collected and preserved in our 

 museums, of which we know only the dead preserved 

 specimens. So that we cannot be sure in doubtful 

 cases whether a series of forms differing a little from 

 the ordinary members of a species indicate distinct 

 species, as defined and tested by breeding. We have 

 in such a case to note the difference, and record it 

 either as a variety or as a species by a guess at the 

 probabilities one way or the other. Naturalists really 

 intend by the word " species " to designate a form 

 represented by numerous like individuals, which, in the 

 present natural conditions of the region they inhabit, have 

 attained a certain "stability" of distinctive form and 

 character (not without some variability within definite 

 limits) and constitute a more or less widely distributed 

 population^ the members of which inter-breed but do 

 not produce offspring with other allied species. 



A good case by which to exhibit further our con- 

 ception of a species is that afforded by the species which 

 are united in the genus Equus the horse-genus. There 

 are living at the present day several wild kinds of Equus 

 namely, the wild horse, or Tarpan, of the Gobi desert of 

 Mongolia, called after the Russian explorer Przewalski; two 

 kinds of Asiatic wild ass, called the Kiang and the Onegar ; 

 the African wild ass, and two or three kinds of zebra. 

 There are, besides, many kinds of domesticated horses, 

 ranging from the Shetland pony to the Flemish dray horse, 

 and from the Shire horse to the Arab. Then there are 



