44 SALMON AND TROUT. 
Another and still simpler attachment for the drop-fly, which 
in practice I usually adopt as being much the quickest, is, 
with a double half-hitch (4 of the knot in fig. 2), to knot on 
the drop-fly—fly uppermost—to the casting line (fig. 5). On 
this knot being pulled tight, and slipped down as far as the 
next juncture on the line, it will be found to answer exceed- 
ingly well, although the point of junction is one which will 
always have to be carefully looked at from time to time, as the 
friction of the drop-fly knot is apt to fray away the link to 
which it is attached. For salmon fishing I never myself use a 
second fly, unless by any chance the river or lake I am fishing 
be also tenanted by white trout, and then, of course, the fly is 
a comparatively small one, for which the last-named attach- 
ment, fig. 5, will answer every purpose ; or slightly better, 
perhaps, the fly may be attached above one of the knots with 
a loop, as shown in fig. 6 ; or, stronger still, as in fig. 7,—an 
attachment which also gives the maximum stand-out-at-right- 
angle inclination to the fly, and the principle of which, as applied 
to casting lines with the ordinary splice, I explained in the 
Modern Practical Angler,’ fig. 8. 
FIG, 6, Fic. 8 FIG. 7. 
Nothing can well be more clumsy than the knots usually 
employed in joining the strands of a salmon casting line, and 
their inefficiency in the matter of strength is on a par with 
their unsightliness. In the ‘ Book of the Pike,’ 1865, I gave 
diagrams and explanations of the buffer knot above referred 
