48 Proceedings of Boyal Society of Edinburgh. [sess. 
estuaries they move to and fro with the flood and ebb tides ; that 
with the ebb they seek the shallower waters at the margins of the 
estuaries, where they are taken in the nets ; hut we need to turn 
to our knowledge of the salmon in fresh water, and the time it 
remains there, in order to form an estimate of how long the fish 
spends in salt water. It seems to me that, even by this process, 
a very uncertain margin is left, hut that in all probability a fish of 
full adult growth, say a fish of four years old, has spent an equal 
period of time in fresh and in salt water. 
It is further noticed that a run of fish commences when the 
river is swollen, or during a spring tide. At such times estuarine 
netting is reported as at its best. 
In the same way, a flood in a particular tributary is believed to 
induce a run of fish, either during the flood or as the water is 
falling after the highest of the freshet is past. If we accept in its 
entirety the theory that salmon return to the particular river where 
they were spawned, we must necessarily believe that the fish which 
take advantage of such a flood are some of those that were spawned 
in the particular tributary ; yet proof is easily found that salmon 
will take advantage of newly accessible spawning grounds, and that 
such fluctuations take place in the conditions of the various tribu- 
taries of a large river as to point to the conclusion that salmon, 
when once in the main river of their birth, enter the particular 
tributaries which are instinctively known to he suitable, or which 
present suitable conditions for ascent at the proper time. What 
these conditions are seem less known to man than to the salmon, 
yet not infrequently fish pass favourite spawning grounds, and, 
impelled it seems by the instinct for ascent, travel on to ground 
where the ova have certainly much less chance of hatching out in 
numbers. 
Last winter I was fortunate enough to have the opportunity of 
procuring data from the head waters of Tweed which, although not 
perfect enough to afford answers to all the questions which sur- 
round the spawning of migratory Salmonidse, give us, nevertheless, 
some information of a more definite kind than can he commonly 
procured, and which on this account are, I think, worthy of 
mention. During the entire spawning season daily records were 
taken by three watchers as to number of fish seen, their position, 
the state of the water, etc. 
