TEMPERATURE OF SCOTTISH LAKES 
95 
The forms of the frame for the reversing thermometer are various, 
one of the forms largely used by the Scottish Lake Survey having 
been designed by the late Mr Parsons (see fig. 35), at one time 
engaged on the Survey, and afterwards principal mineral surveyor in 
Ceylon ; but the essence of all the frames is that on a spring catch 
being released the thermometer is inverted. In the usual form of 
frame the spring is released by allowing a weight, called a messenger, 
to slide down the sounding-line till it strikes the thermometer frame. 
The form of messenger to which I am most partial is a bar of metal 
twisted into a spiral, which can then be wound on to or off the 
sounding-line without fear of loss (see fig. 36). This form of 
messenger was designed by Mr William Macdonald, my boatman and 
observer on Loch Garry. By means of these thermometers observa- 
tions may be rapidly taken, the rapidity being increased by attaching 
thermometers to the sounding-line at intervals ; two or three or more 
thermometers may be used on one line at the 
same time, and thus simultaneous observations 
of temperature at the several depths may be 
obtained. The thermometers had, of course, 
to be immersed for some time to allow of their 
taking up the temperature of the water sur- 
rounding them ; and it was found by experi- 
ment that, where there were no great differences 
of temperature in the course of the observa- 
tions, three minutes was sufficient time to allow. 
The early observers worked under great 
disadvantages, having to use an ordinary 
thermometer in one of two ways. In one method a thermometer, 
with many coats of non-conducting material, was lowered to the 
desired depth, and left for many hours to take up the temperature 
of the surrounding water. The protective coatings prevented 
any appreciable alteration in the temperature while being raised 
to the surface, and the thermometer could then be read. The 
other method was to bring up a sample of water from the desired 
depth in a suitable form of bottle or bucket, and then measure the 
temperature of the water when it reached the surface. It will be at 
once apparent that these methods were far from perfect, but in the 
hands of observers like Saussure they were used with great accuracy, 
and the observations may be relied on implicitly. 
Subsequently many varieties of self-registering thermometers 
adapted for observing deep-water temperatures were designed, but 
in connection with these it is only necessary to mention the names 
of Cavendish, Six, and Aime. The Millar-Casella maximum and 
minimum thermometer, which is a form of Six's thermometer, is 
