322 THE INFLUENCE OF INANIMATE SURROUNDINGS. 



contrivances must exist which neutralise the ill effects of the 

 compression of the contained air which must undoubtedly take 

 place. 



In the instances here adduced, and in other similar ones, of 

 the action of gravitation on animals, the effects are obviously 

 merely selective ; all the individuals which are not quali6ed to 

 accommodate themselves to the actual conditions of pressure must 

 perish or seek a more suitable habitat. But gravitation may 

 perhaps have also a direct determining action, perhaps in a 

 mode analogous to that by which the growth of the roots of 

 plants, or the structure of the underside of leaves, and other 

 things may directly depend on gravitation. The theoretical 

 possibility of this influence is beyond dispute ; but we know 

 very little of its actual effects and extent. Nor can there be 

 any doubt that animals, in consequence of their greater freedom 

 of movement, are in a great degree independent of it ; and any 

 extensive influence of this kind, such as is undoubtedly mani- 

 fested in plants, must be out of the question, except as regards 

 sedentaiy animals, such as corals, sponges, bryozoa, &c. How 

 far, in such creatures as these, gravitation may have an effect in 

 determining the general form of the colony, or of the individual 

 animals and their organs, is perfectly unknown, and it is diffi- 

 cult to see by what means it would be possible to ascertain 

 experimentally the effects of gravity upon such animals. For 

 all those contrivances which have been successfully employed 

 on plants, to allow gravitation to exert a perfectly independent 

 influence on their growth, cannot be applied to animals, and, 

 so far as can be seen, we can only fall back on the interpre- 

 tation of those experiments which Nature herself performs on 

 growing animals under the normal conditions of their existence. 

 It is evident that we can thus only arrive at more or less bold 

 or plausible hypotheses ; for the fact cannot be too often insisted 

 on that experiment alone can ever enable us to explain the 

 causes lying at the root of any particular phenomenon in the 

 development of an animal. All theories deduced only from the 

 visible phenomena without the counter-check of experiment 

 are mere clever suggestions, which only serve to conceal our 

 ignorance, and in fact hinder any advance. Thus, for instance, 



