ABOUT HOOKS. 241 



selves intelligible to the reader we shall first describe the 

 usual system of making the hook. First, then, the wire 

 is struck off in given lengths, in accordance with the size 

 of the hook required ; next, the point is formed and the 

 shank reduced by a few strokes of the file ; and next, the 

 barb is cut by means of a large knife. All is now ready 

 for bending, which is one of the most particular items in 

 the construction, as the operation decides the shape, and 

 consequently, the particular species of hook to be pro- 

 duced. This is quickly done by means of a small steel 

 block around which the wire is bent, the shape of the 

 block varying according to the particular bend required. 

 Now comes the final operation, viz., that of tempering. 

 This is done in a large pan over a slow furnace. Millions 

 of books are frequently tempered in one operation, there- 

 fore the greatest care should be bestowed upon this im- 

 portant point ; but of this more anon. The most fertile 

 source of complaint is, we believe, the undue weakness 

 of the majority of hooks at the barb (see diagram VI. , 

 fig. 2). A deeply barbed hook may be safely discarded 

 as being too dangerous to use. There is no earthly ne- 

 cessity for it. Not only is the wire half cut through by 

 the operation, but the point of the hook is forced quite 

 out of the straight line ; thus, not only requiring heavier 

 striking to hook a fish properly, but being more liable to 

 snap by the sudden strain. The very action of inserting 

 the bearding knife too deeply causes the point to project 

 outwards inadvertently (see Plate 6, fig. 2), giving it a 

 "scratch and let go" appearance, and no less effect ; in- 

 deed, there are but two alternatives for a hook of this 

 description when a fish is caught upon the gristly or bony 

 part of the mouth (1) to scratch as described, or (2) to 

 break clean off at the barb near the point of the hook. 

 We have seen hooks fitted with costly salmon flies, nine- 

 tenths of them being quite useless, having broken at this 

 identical point. A batch was shown to us by (the late) 

 11 



