GEOGRAPHICAL DISTRIBUTION I49 



Small has shown that the local species have regions of 

 concentration along the paths of migration of the wide- 

 spread species, and that they are most abundant "along 

 the ridge which extends around the Pacific and Indian 

 Oceans from Tierra del Fuego to South Africa." The 

 paths of migration are chiefly coextensive with the alti- 

 tude of 3,000 ft., or higher, and all the facts point to the 

 Andes of Bolivia as the probable (h)^othetical) center of 

 distribution for the genus, whence it has rapidly spread 

 " along the unwooded regions of the mountain ranges of the 

 world." This world-wide distribution, and the posses- 

 sion of pappose fruits which would make possible a wide 

 distribution in a relatively short period of time (geologi- 

 cally speaking), all point (as do the facts of its morphology) 

 to the comparative youth of the group; while its marked 

 tendency to variation, its success in the struggle for exis- 

 tence (as may be noted everywhere), and finally the- 

 existence of innumerable local species, with centers of 

 distribution along the paths of migration of the genus as a 

 whole, are just the facts which one would expect to find 

 on the basis of the theory of evolution. 



115. Dispersal by Water and Birds. — Space at our 

 disposal will permit of only a passing reference to seed- 

 dispersal by water and birds. In order to be carried for 

 long distances by water, seeds and spores must^be able to 

 undergo prolonged soaking in water, and in the case of 

 ocean currents, in salt water. Many species of the new 

 strand flora of Krakatoa were certainly brought many 

 miles by ocean currents, and Guppy, who made a study 

 of "Plants, seeds, and currents in West Indies and Azores,"^ 

 cites the case of a ragweed {Ambrosia crithmifolia) whose 



' Guppy, H. B. London, 1917. 



