462 
Proceedings of the Royal Society 
the east : its prolongation forms a bar across the channel with not 
more than 2 feet of water on it. Since the bar and hanks are con- 
stantly shifting in size, shape, and position, all observations which 
depend in any way on the form of the river mouth must be con- 
sidered as applicable only to the time when they were made. 
The salmon-fishing station of Tugnet stands on the east shore 
about one-third of a mile from the river mouth. A backwater 
separated, except at high tide, from the main stream by a sandy 
flat, runs up from near the river mouth to Tugnet. On the west 
side a narrow shallow channel, separated by a low grassy island 
from the river, comes down past Garmouth, and at its termination a 
burn which trickles across the mud of Kingston Basin at low tide, 
enters the river. At high tide this burn becomes a passage with 
considerable depth of water, leading to a broad expanse below the 
houses at Kingston, and separated from the sea by a high shingle 
bank. The relative positions of these parts of the river may be 
understood from the accompanying sketch (fig. 1), which is drawn 
roughly from memory, and is not to scale. 
Even when the river is small and the tide full the water is fresh 
on the surface.* The rising tide diminishes the velocity of the 
current, both by damming back the fresh water and by widening the 
connection with the sea, and so letting it spread over a larger area. 
The conditions of working in a small coble in a rapid stream of 
slight depth, in which considerable variations of salinity might be 
expected, necessitated some modifications of the usual modes of 
obtaining and examining samples of water. The slip water-bottle 
was useless as a collector, because it enclosed a vertical column of 
water a foot high, and it soon became apparent that very consider- 
able changes of density took place in a distance of 3 inches. It 
was found best in the circumstances to use an ordinary stoppered 
bottle (of about 1*5 litre capacity) lashed to a pole or attached to a 
sounding-line which was heavily weighted, t and to draw out the 
* The fact that sea water forced its way up the bed of a tidal river under 
the opposing current of fresh water has been known at any rate since 1812, 
when the phenomenon was studied on the Dee at Aberdeen by Mr Robert 
Stevenson. — David Stevenson, Canal and River Engineering, 2nd ed., p. 124. 
+ The hydrophore of Mr Robert Stevenson ( Canal and River Engineering, 
p. 126) would have been more suitable, but there was no opportunity of getting 
one made. 
