NA TURAL HISTORY OF BRITISH SALMONIDA. 131 
the instances quoted is far from being invariable ; and as it 
has been found that the time of salmon ascending and spawn- 
ing frequently differs in neighbouring rivers of the same district 
—in some cases even where their sources and channels are 
apparently of a similar nature—it is very possible that we have 
yet to arrive at the whole truth respecting the causes of these 
variations. 
With regard to the sort of resting places or holding grounds 
which salmon most frequent, they appear, as often as not, to be 
guided by sheer caprice. There are some pools which, to the 
angling eye, or, at any rate, the uneducated angling eye, are 
apparently perfection, and which yet seem never to hold a 
fish from year’s end to year’s end, orif they do holda fish 
they are fish which cannot be induced to take the fly or bait. 
Indeed, there are particular stones in particular rivers behind 
which for some inscrutable reason, salmon will almost invari- 
ably be found. I know just such an one in the Conway, and 
there is another, if I recollect rightly, in the Tweed. It does 
not signify how often the salmon sheltering behind these stones 
is caught—I have known three thus taken in one day—his 
place appears to be filled again almost immediately. 
Still, notwithstanding this capriciousness of salmon in the 
choice of water, there are some general rules which may serve 
as approximate guides to the salmon fly-fisher when unassisted 
by local knowledge. Where the bed of the river is of bare or 
naked rock, unbroken by ledges or ‘shelters’ of any sort, 
salmon—or, at any rate, rising salmon—will rarely be found. 
All sorts of shelters and rocky excrescences are, no doubt, in 
themselves favourable for salmon, both as affording shelter 
from the stream and a point of outlook from which fly and other 
bait may be advantageously perceived ; but, as Mr. Stoddart 
also points out in his ‘ Angler’s Rambles,’ the ledges of rock 
and large stones to which salmon instinct inclines ‘will in- 
variably be found, when the salmon are settled down, to lie 
in conjunction with or in the vicinity of a firm gravelly 
“ alveus.”’ Elsewhere, Mr. Stoddart illustrates this fact by the 
K2 
