AND GEOLOGICAL RELATIONS OF CERTAIN HOT SPRINGS. 
577 
The first point we have to attend to is the true scale of the thermometers employed. 
§ 1. Verification of Thermometers. 
In by far the majority of the experiments to be detailed one thermometer was em- 
ployed. It is a standard thermometer made for me by Troughton and Simms, pur- 
posely for these experiments, in 1832. In my experiments in the Pyrenees, when 
temperatures below 125° Fahr. occurred, another thermometer (a pocket one by 
Crichton, sen., of Glasgow,) was generally employed simultaneously, both as a 
check upon the other observations, and in event of Troughton’s being unfortunately 
broken. Troughton’s instrument is happily entire, and to it reference will always be 
made when no other is specified. Some observations taken by Crichton’s thermo- 
meter alone, will be reduced to a true scale through the medium of Troughton’s, 
with which, as will be seen, I have preserved abundant comparisons. 
In all the experiments made in France with Troughton’s thermometer, the fol- 
lowing precautions were rigidly observed. 
1. The scale was carefully immersed in the water up to the point indicated by the 
mercurial column. 
2. The thermometer was held in a horizontal position, or nearly so, when practi- 
cable ; or if not, the position was specified. This was owing to a suspicion, that in 
consequence of the pressure of the column of mercury on the bulb, the indications 
were lower in a vertical than in a horizontal position. Since my return I put this to 
a careful proof as follows : The bulb was placed in a tube of stiff paper, so as to pro- 
tect it from external pressure. The thermometer so defended was placed in the axis of 
a tin cylinder filled up with sand, the upper part of the scale projecting. The whole 
was gradually heated till the thermometer indicated above 200°, and varied very 
slowly, owing to the difficultly-conducting envelope. Being then observed many 
times in succession in a horizontal and vertical position, I found the excess in the 
former extremely small, not exceeding 0°‘15 even at that high temperature. I have 
consequently considered any error on this score as negligible. 
3. The temperature of different parts of a stream or reservoir has been noticed, 
and if any difference occurred, the mode of observation described. 
4. The observations have been made always to tenths of a degree of Fahrenheit 
by estimation ; and all are in that scale unless otherwise expressly mentioned. 
But by far the most important precaution was the ascertainment of the true scale 
of the instrument, an element which I did not consider myself entitled to assume as 
yet diminished extremely the discharge of the hot springs. When, in consequence of legal measures, the pro- 
prietor was forced to close the aperture he had made, the water returned to the original springs in the same 
quantity and having the same temperature as before. 
P.S. January 1837. Since the above was written, experiments have been made by M. Freycinet on the 
springs at Aix, by direction of the Academy of Sciences. See the Comptes -Rendus des Seances de l’Academie, 
1836, l er Semestre. 
4 E 
MDCCCXXXVI. 
