106 



Tlie American Salmon-fishernuMi, 



Fig. 8. 



I then construct in imagination the parallelogram h c 

 de, of which the barh-side of the point, a, is the diagonal, 

 as sho\m in the following figure. 



Kg. 9. 



Tlien, unless the side h c considerably exceeds in length 

 the perpendicular c e, the hook is condemned. It is 

 obvious that the barb shown on the left-hand side of the 

 figure is much more likely to rake its way out of a fish's 

 mouth than that on the right hand, while the latter will 

 bury sufficiently to insure a good hold if it buries at aU. 

 Therefore in certainty the right-hand barb is superior, 

 while practically equal otherwise. The shorter the side 

 b c, when compared to the side c e, the more defective is 

 the hook in this respect, since when the side b c becomes 

 nothing there is no tendency to penetrate at all. The 

 relation of these sides of the parallelogram to one another 

 expresses the certainty of the hook to bite; and there- 

 fore the rule as given above. 



