of Edinhunjh, Session 1883-84. 689 
sunshade. The sheath was found to act almost as well as the 
chemically deposited silver, hut not quite, its readings being gene- 
rally about 0°*2 below the fan thermometer. As silver has a tendency 
to tarnish, it was thought as well to try the efficiency of gold as a 
protector. Gold is not quite so well suited for the purpose, as it 
absorbs more radiant heat than silver ; while silver only absorbs 
3 per cent., gold absorbs 5 per cent. The first sheath which was 
prepared for the thermometer, and which was ipade of impure silver, 
was gilt with gold, and tested under the same conditions as the pure 
silver sheath. It was found to be not nearly so efficient a protector 
as the silver, its readings being generally about 0°*2 above the 
fan thermometer, or nearly 0°*4 higher than the pure silver. The 
gold gilt sheath, after being very much finger-marked, made the 
thermometer read nearly 0°‘5 above fan, or about half a degree 
better than clean glass. As to whether gold or silver is the best 
for practical purposes can be determined only by continued use. 
Gold has certain advantages, as it will probably keep its protecting 
powers longer than silver ; and where frequent cleaning is incon- 
venient, gold may be the most suitable; but when great accuracy is 
aimed at, silver must be employed. 
The action of these silver-coated thermometers is most curious 
and interesting. They never agree with any other thermometers 
when hung alongside of each other in a room into which the sun is 
shining, or in which the gas is lighted ; and even after the windows 
are closed, and shutters shut, or the gas put out, it is long before 
they agree with the others. Like captious critics, they always under- 
estimate the radiations emanating from surrounding bodies, and 
manage to keep themselves cool amidst the heat exchanges taking 
place on every side. If we examine the readings of an ordinary 
thermometer, it is influenced by our presence; and if long in making 
our observation, the heat radiated from our body may cause the 
thermometer to indicate a temperature half a degree too high ; but 
these silvered thermometers are almost indifferent to our presence, 
while they are sensitive to the temperature of the air. 
Sling Thermometers. 
The sling or “ Fronde ” method of taking observations has 
generally been accepted as the most accurate way of determining 
2 y 
VOL. XIT. 
