21 8 THE FORMATION 



involucre that the latter serves as a sort of mortar for projection, when the 

 stem of the plant is bent to one side by any force, such as the wind or an 

 animal. It will be noticed that two separate agents are actually concerned in 

 dissemination of this sort. 



Frequently, two or more agents will act upon the same disseminule, 

 usually in succession. The possibility of such combinations in nature is 

 large, but actual cases seem to be infrequent, except where the activities of 

 man enter into the question. Some parts, moreover, such as awned inflor- 

 escences, are carried almost equally well by wind or animals, and may often 

 be disseminated by the cooperation of these two agents. The wind also 

 often blows seeds and fruits into streams by which they are carried away, 

 but here again, parts adapted to wind-dissemination are injured as a rule 

 by immersion in water, and the number of plants capable of being scattered 

 by the successive action of wind and water is small. 



In the present state of our knowledge of migration, it is impossible to 

 establish any definite correspondence between dissemination-contrivance, 

 agent, and habitat. As a general rule, plants growing in or near the water, 

 in so far as they are modified for this purpose at all, are adapted to water- 

 carriage. Species which grow in exposed grassy or barren- habitats are for 

 the most part anemochores, while those that are found in the shelter of 

 forests and thickets are usually zoochorous, though the taller trees and 

 shrubs, being exposed to the upper air currents, are generally wind-distrib- 

 uted. There is then a fair degree of correspondence, inasmuch as most 

 hydrophytes are hydrochorous, most hylophytes, zoochorous, and the ma- 

 jority of poophytes and xerophytes, anemochorous. Definite conclusions 

 can be reached, however, only by the statistical study of representative 

 formations. 



With respect to their activity, agents may be distinguished as constant, 

 as in the case of currents, streams, winds, slope' growth, and propulsion, 

 or intermittent, animals and man. In the former, the direction is more or 

 less determinate, and migration takes place year by year, i. e., it is contin- 

 uous, while in the latter dissemination is largely an accidental aflfair, inde- 

 terminate in direction, and recurring only at indefinite intervals. The 

 effective conversion of migration into invasion is greatest when the move- 

 ment is continuous, and least when it is discontinuous, since, in the latter, 

 species are usually carried not only out of their particular habitat but even 

 far beyond their geographical area, and the migration, instead of being 

 an annual one with the possibility of gradual adjustment, may not recur for 

 several years, or may, indeed, never take place again. The rapidity of 

 migration is greatest in the case of intermittent agents, while the distance 

 of migration is variable, being great chiefly in the case of man, ocean-cur- 



