306 SALMON AND TROUT. 
The first and most obvious method, then, for counteracting 
the causes to which I have pointed as tending to reduce the 
volume of our streams and the amount of trout food which 
they supply, lies in deepening and widening portions of those 
streams. This can be easily done in many of our brooks, by 
raising barriers to hold up the water, and by enlarging and 
deepening portions of their courses at the small sacrifice of a few 
square yards of poor soil adjoining a natural hollow in their beds. 
The fish in the artificial pools thus formed will be better fed and 
consequently larger than those in the ordinary shallow course 
of the brook or ‘ pelting river’—to borrow Shakespeare’s phrase 
—which favours the multiplication of trout but fails to supply 
them with abundant food. 
Of course we must remember that trout water, whether pool 
or river, may easily be overstocked. In the course of a 
ramble through an unfrequented part of Lochaber, I once came 
upon a tiny tarn, fed by a burn which, though of the smallest 
size, afforded excellent gravelly bottom for ‘redds.’ I madea 
few experimental throws over it, and each time landed a fish 
on every fly. I added two small hackles to my ordinary cast 
of three, and had five troutlings hooked in as many seconds. 
I made a dozen more casts, and each time took five fish. They 
were so greedy that they wou/d have the hook, so small that I 
had no difficulty in sending the whole quintett flying. Had I 
had any object in further slaughter—a feud with the cook at 
Inverlair, or an extensive’ contract for potted trout—I could 
easily, with the aid of my gillie to unhook the fish, have takena 
thousand brace of these hungry fry in a day. Mine were 
perhaps the first artificial flies they had ever seen, for the tarn 
in question lies quite off the beaten track, though near Lochs 
Treig and Ouchan, which would have naturally attracted any 
wandering angler in those regions. But such a case of over- 
stocking I never witnessed. 
Within a mile or two, and on the same stretch of moorland, 
but at a lower level and where the depth of peat was far greater, 
lay another tarn of four or five acres in extent, which had no 
