THROWING THE FLY. 103 



laws would necessarily operate more or less if the 

 spring were made in any part of the arc o b, 

 except at b itself, and becoming less and less 

 objectionable as the spot of the spring approached 

 to B. Again, if the spring backwards were made any 

 where between b and h, the nearer it approached 

 H the greater would be the impossibility of even 

 raising the fly from the water, and the possibility 

 of throwing the line back to the water would only 

 be superinduced as the point of spring approached 

 nearer to b. The reason why the forward spring 

 is made nearer the perpendicular than the back- 

 ward one, is that the fly is then allowed to ap- 

 proach less near to the ground in passing from e 

 to J, and there is consequently l^ess danger of its 

 catching. It is also thrown forward nearer to 

 a parallel with the earth, and then its descent on 

 the water will be so much the lighter than if 

 it were thrown more up towards n, and thence 

 fell to M, or if thrown more in a straight line with 

 M, and so struck the water instead of gradually 

 alighting upon it: while, by making the backward 

 spring further from the perpendicular, the fly is 

 raised higher in the air, and traces a parabola, at 

 the termination of which the fly is caught by 

 the forward spring before it approaches within 

 a "catching" distance of the ground behind. So 

 much for the science of plain throwing, in which 



