If the chemistry of mineral waters has been as yet prosecuted to a very limited extent, notwithstanding the number of eminent analysts who have engaged in the research, much more has every other topic connected with their origin and nature been superficially treated. The characters of springs of every kind are so important as to deserve minute and laborious research; and notwithstanding the partial essays of Von Buch and others, the whole subject remains in a state of confusion, and is involved in incongruities and contradictions. The chief point to which the observations about to be described were directed, is the temperature of thermal springs; and, referring to this point alone, we might prove the almost total absence of exact data on the subject. Every traveller, to be sure, has measured the temperature of springs, but few have been aware of the difficulties which even this apparently simple inquiry involves. We should have considered the accurate determination of the temperatures of thermal springs the first step towards a theory of their production. The constancy of that temperature from day to day, from year to year, from century to century, would appear one of the most essential facts to determine; yet I am aware of scarcely a single published observation capable of being satisfactorily employed in such an inquiry. Not only are the errors of the instruments unknown, but the circumstances under which the observations are made are liable to perpetual change. For instance, a spring in the state of nature may rise from rock directly, or from amongst debris. In the latter case, to fix the temperature is difficult, because it varies at different points; and it is nearly useless, because a year hence the circumstaqces of its efflux may be wholly changed. Again, in the more usual case of thermal waters being medicinally employed, it is frequently impossible (at least without much trouble) to reach the true source of the water, which is carried through pipes, conduits, and reservoirs before it is finally employed; and in this case the temperature is usually taken at the bath-cock, or at the 'buvette,’ or drinking-cock, where consequently the water has been subjected to the variable cooling action of its intermediate transit. Thus, for example, at the great establishment of La Raillière at Cauteretz, in the Hautes Pyrénées, the water is cooled from 101°·9 to 99°·8 in passing through a short and well-inclosed stone conduit from the source to the 'buvette’; and in the neighbouring spring of the Mahourat, the spout from which the water flows, though in contact with the granite rock from which it rises, and, in common parlance, the true or real source, I found to give a temperature 0°·5 lower than I obtained a few feet further back by squeezing myself into an almost inaccessible cleft of the rock. Thus for the most part we have no assurance that two travellers have observed the same spring at the same point; and hence identity of name by no means infers comparability, even supposing the instruments perfect. The frequent alterations in the thermal establishments render a specific description of the locality still more indispensable. Where the mineral water is not applied to use, we have a new difficulty in the recognition of a spring by the mere description of locality. That hot springs should ever be so abundant as to render this possible might seem improbable; I have had occasion to suffer from it, however, in following the footsteps of the indefatigable Anglada amongst the numerous and often almost inaccessible hot-springs of the Eastern Pyrenees (near Thuez, in Roussillon).
XXVI. On the temperatures and geological relations of certain hot springs, particularly those of the Pyrenees; and on the of verification of thermometers
