FOOD OF FISH AND ITS PRODUCTION. 101 



the larvse of the insects whicli will almost always 

 be found to abound in them, from one lake and from 

 one stream to another ? For example, with respect to 

 the scjuillidse so often noted, what could be easier 

 than to transplant weed ? This would serve as food 

 for the large fresh-water muscle found in almost all 

 waters, and it would serve as food for the squillidae, 

 which in turn would serve as food for the fishes. It 

 may be said, vrith regard to some lakes and streams, 

 that they are so gravelly or rocky, that the weeds 

 would h&rdly thrive in them ; but it is seldom indeed 

 that some nooks and corners do not exist, in or about 

 the banks of lakes and streams, where there may be 

 found sufficient soil, which, with a slight admixture 

 of the natural soil, and a judicious planting of these 

 weeds, may not be made to grow them to some small 

 extent; and the weeds, once introduced, wUl gradually 

 increase year by year, forming their own soil, and 

 naturally producing those requisites which are the 

 most favourable to their production. Of course judg- 

 ment must be exercised in carrying out such experi- 

 ments, quite as much as would be exercised in the 

 introduction or cultivation of a new food-producing 

 plant in agriculture. We acclimatize every species 

 of agricultural plant, and examine its qualities and 

 capabilities, for cattle or for ourselves ; we study 



