DISPERSAL 



159 



find their proper habitat, when they are migrating or otherwise 

 becoming dispersed, there are three important devices to 

 be noticed. The first class of animals employs the method of 

 broadcasting enormous numbers of young ones or even of eggs, 

 with the result that some of them fall on stony ground and 

 perish, while others reach the right place, where they have at 

 least a very thin chance of establishing themselves. This 

 method is employed by all those spiders which float away on 

 gossamer threads when young ; by locusts which start on migra- 

 tion with full air-sacs and loaded fat-bodies, and fly along until 

 the air and the fat are used up, when they have to descend and 

 can then only undertake small local movements ^^ ; and it is 

 also used by an enormous number of sessile and sedentary 

 marine animals, which produce free-floating larvae that have 

 only small powers of directive movement. This broadcasting 

 involves a huge wastage of life, and is usually confined to young 

 animals, except in cases which, like the locusts, are partly 

 concerned with the reUef of pressure in the home population. 



The second method is to have some special reaction which 

 enables the animals to find their suitable habitat ; this is a 

 very widespread method, and is much less wasteful than the 

 broadcasting one. Pettersson ^^^ has studied the fluctuations in 

 the Baltic herring fisheries and found that the herring pro- 

 bably only enter the Baltic when water of a certain salinity 

 penetrates there. They follow the salt water, and refuse to 

 go in water with a salinity of less than 32 or 34 per mille. 

 At the present time the lower layers of the Baltic are 

 hardly ever more than 28 per mille ; but this condition 

 is influenced by the tidal effects of the moon, so that it 

 appears that in the past there have been at certain times 

 invasions of the Baltic by comparatively salt water, such as 

 still occurs in the Skagerak and Kattegat. His work on this 

 problem makes it practically certain that the existence of a 

 definite salinity-preference on the part of the herring, combined 

 with peculiar tidal phenomena, has caused in the past regular 

 fluctuations in the prosperity of the Baltic and neighbouring 

 fisheries, whose period is about i8| years, with a superimposed 

 longer period of iii years. 



