158 
Proceedings of the Royal Society 
considered advisable to make the results public. This is done not 
only to suggest the general law governing the distribution of tem- 
perature in an estuary at different seasons, but to get the benefit of 
any suggestions that may be made as 
to the carrying on of the work in 
future. 
The methods of observing the 
temperature of water are extremely 
simple. Tor the surface it is sufficient 
to place a thermometer for a few 
seconds — the time must be greater 
the more sluggish the instrument — - 
in a bucket of the water, or to dip 
the thermometer in the water and 
read it rapidly on its withdrawal. 
In using the latter method, the ther- 
mometer case must terminate in a 
little cup to contain enough water to 
cover the bulb. Tor depths below 
the surface the hTegretti and Zarnbra 
thermometer in the Scottish frame * 
has been found most suitable. 
A special arrangement requires to 
be made in order to fix continuously 
immersed self-registering thermo- 
meters at the surface and bottom, so 
that they will not have their distance 
from the bottom and surface altered 
by the rise and fall of the tide. This 
was accomplished at Granton by 
banging the surface thermometer 
B (fig. 1), an ordinary Miller-Casella 
self-registering maximum and mini- 
mum instrument, to an anchored buoy 
A. The bottom thermometer was 
also attached to the buoy with a line long enough to reach the 
bottom at the highest spring tides. In order to prevent the ther- 
* Proc. Roy. Soc. Edin xii. p. 927. 
