150 Mr Sang on the relation which subsists 



still more so. As we approach the larger quadrupeds, orna- 

 ment, and then elegance, disappear. This is the law in the 

 works of Nature, and this ought to be the law among the works 

 of Art. 



Among one class of animals, indeed, it may be said that this 

 law is reversed. We have by no means a general classification 

 of the fishes ; but, among those with which we are acquainted,^ 

 we do not per<;eive such a prodigious change of form. Here, 

 however, the animal has not to support its own weight; and 

 whatever increase may take place in the size of the animal, a 

 like increase takes place in the buoyancy of the fluid in which 

 it swims. Many of the smaller aquatic animals exhibit the ut- 

 most simplicity of structure ; but we know too little of the na- 

 ture of their functions to draw any useful conclusions from this 

 fact. 



Having said thus much on the relative strengths of a machine 

 and of its model when at rest, I proceed to compare their 

 strengths and actions when in motion. 



This subject naturally divides itself into two heads ; the one 

 relating to the ability of the structure to resist the blows given 

 by the moving parts, either in their ordinary action, or when, 

 by accident, they escape from their usual course ; the second 

 treating on the changes which take place on the friction of the 

 parts when these are enlarged or diminished. 



The ability of a support to resist the impetus of a moving 

 body, is estimated by combining the pressure which it is able to 

 bear with the distance through which it can yield ere disruption 

 take place. In the case of a support which acts longitudinally, 

 the strength is proportional to the square of the linear dimension, 

 while the distance through which it can yield is as the linear di- 

 mension itself. Altogether, then, the ability to resist a blow is 

 proportional to the cube of the length ; that is, to the weight of 

 the body which is destined to act upon it. If, then, the linear 

 velocity of the machine is to be the same with that of the mo- 

 del, these parts, so far as this action is concerned, will be in 

 keeping with each other. 



' In the case, however, of a lateral support, the distance through 

 which it can yield without breaking is not augmented by an en- 



