216 THE ORIGIN OF SPECIES [pt. ii 



upon the theory of mutation. We have no evidence to show that 

 a Jordanian species will proceed further towards a Linnean 

 species. One cannot imagine the 11 Doo?ias, or the 15 species of 

 Stemonoporus in Ceylon (p. 152) arising in this way. The 

 Jordanian varieties show the same phenomena of dispersal as 

 do the Linnean species of Ceylon and elsewhere, and often 

 occupy as large areas, while they still remain true-breeding, and 

 show no sign of variation. As a general rule, it is not hard to 

 place a Jordanian species in its proper "Linnean" aggregate. 



In Aaew of the large mutations that have been recorded, e.g. 

 CapseUa Heegeri (104) and others (cf. list in 39, p. 308), and 

 upon general grounds of comparison of the characters used in 

 systematic work upon the classification of the flowering plants, 

 it seems to me that mutations may at times be of the necessary 

 size to give rise at once to Linnean species. One cannot con- 

 ceive of the many species of Ranunculus in New Zealand, for 

 example, arising by the gradual separation of Jordanian varieties, 

 especially when these breed true. ^Ye have no evidence to 

 show that the intermediate forms, as would be necessary on this 

 hypothesis, die out. The struggle for existence comes at the 

 moment of birth of a species, and if it survives it may spread. 



The view that mutations are necessarily small rests upon the 

 opinion, often put forward as if it were a general rule {e.g. 67), 

 that a Linnean species consists of a great assemblage of micro- 

 species, which breed true, as has been shown to be the case in 

 Erophila verna, for example. But this opinion requires a com- 

 plete re\'ision in \\e^y of the facts that have been set forth above 

 in regard to Age and Area. A species can only consist of such 

 an assemblage, obviously, if it consist of many individuals, and 

 occupy a large area of ground. Now in the north temperate zone, 

 where most of our botanical research is carried on, this is in fact 

 true of nearly all species; and only a few are localised, for in- 

 stance in the Alps or the Rockies, or to a less extent in the 

 plains, particularly of North America and West Asia. These 

 localised species have been looked upon as relics or special local 

 adaptations, and often disregarded from an evolutionary point 

 of view. But the work that has been done upon Age and Area 

 shows that such species, except to some extent within the range 

 of the effects of the glacial periods, must be regarded as young 

 beginners. Now in their case, where often the Avhole species is 

 only represented by a few individuals, it is clear that unless every 

 plant or two is different in hereditary characters, the species 



