228 SALMON AND TROUT. 
curve of which is facing down stream, is changed in its direc-. 
tion to DG; the outward curve facing up stream, the posi- 
tion of the rod being shifted from c top the action of the 
stream will then straighten the line, which will gradually get 
into the position cH, the position of the rod being shifted back 
to ac; the fly will then work gradually across stream, the rod 
following the direction of the fly until the cast is completed at 
ABM. Few fishermen IJ have watched fishing take the trouble 
to take the belly out of their line, and are content to let the fly 
work in the same position as it was cast; but if they would 
look at it in the light I see it, I feel convinced they would be 
of my opinion. 
Many experienced fishermen advocate casting straight across 
stream, and assert that by adopting this method many more 
fish are risen than by any other ; they may be right, but I much 
doubt it, and maintain that, even should more fish be risen by 
the straight-across method, more fish are killed by casting 
down and across. 
A reference to diagram, fig. 4, will show how a fly works 
cast straight across stream, A BC, from the time the fly touches 
the water at c to when the cast is completed, ABE. The course 
of the fly is represented by the dotted line cD; the position 
of the rod cannot be altered, as it would make matters worse. 
It will be seen that the fly is travelling from first to last head 
foremost down stream, the cross action of the stream on the 
fly will put all its feathers out of gear, the fish in the pool will 
-get but a momentary sight of it, and will have no time to dis- 
cern its colour, and if they rise at it, by the time they reach 
the surface of the water the fly will be ayard behind them down 
stream, and the disappointment thus caused will be apt to scare 
them to such a degree that they will not rise again. 
That fish are thus caught I do not deny, but I maintain 
that many more are caught by adopting the down and across 
cast. 
Figs. 5 and 6 are diagrams representing the two slovenly 
casts I have before alluded to. In both diagrams Bc shows 
