MIGRATION 49 



the enormously greater number of disseminules, and their 

 much greater chance of becoming established. 



The direction of migration is indeterminate, except in the 

 case of those distributive agents w-hich act constantly in the 

 same direction. , The general [tendency is, of course, for- 

 ward, the lines of movement radiating in all directions from 

 the parent area. This is well illustrated by the operation 

 of winds which blow from any quarter. lu the case of the 

 constant winds, migration takes a more or less definite 

 direction, the latter being determined to a large degree by 

 the fruiting period of any particular species. In this con- 

 nection, it must be kept clearly in mind that the position of 

 new areas with reference to the original home of a species 

 does not necessarily indicate the direction of migration, as 

 the disseminules may have been carried to numerous other 

 places in which ecesis was impossible. The local distri- 

 bution of zoochorous species is of necessity indeterminate, 

 though distant migration follows the pathways of misrratory 

 birds and animals. In so far as dissemination by man takes 

 place along great commercial routes, or along highways, it 

 is determinate. In ponds, lakes and other bodies of stand- 

 ing water, migration may occur in all directions, but in 

 ocean currents, streams, etc., the movement is determinate, 

 except in the case of motile species. The dissemination of 

 plants by slopes, glaciers, etc., is local and definite, while 

 propulsion is in the highest degree indeterminate. Migra- 

 tion by growth is equally indefinite, with the exception that 

 hydrotropism and chemotropism result in a radiate move- 

 ment away from the mass, while propulsion throws seeds in- 

 differently into or away from the species-mass. From the 

 above, it will be seen that distant migration may take place 

 by means of water, wind, animals or man, and, since all 

 these agents act in a more or less definite direction over 

 great distances, that it will be in some degree determinate. 

 On the other hand, local migration will as regularly be in- 

 determinate, except in the case of streams and slopes. The 

 direction of migration then is controlled by these distribu- 

 tive agents, and the limit of migration is determined by the 



