2. Mineral waters have at all times attracted the attention of the faculty in consequence of their peculiar properties and medical virtues. Some faint steps towards their investigation were taken by Boyle. Du Clos attempted a chemical analysis of the mineral waters in France; and Hierne made a similar investigation of the mineral waters of Sweden. Though these experiments were rude and inaccurate, they led to the knowledge of several facts respecting mineral waters which chemists were unable to explain. One of these was the existence of a considerable quantity of _calcareous earth_ in some mineral waters, which was precipitated by boiling. Nobody could conceive in what way this insoluble substance (_carbonate of lime_) was held in solution, nor why it was thrown down when the water was raised to a boiling heat. It was to determine this point that Mr. Cavendish made his experiments on Rathbone-place water, which were published in the year 1767, and which may be considered as the first analysis of a mineral water that possessed tolerable accuracy. Rathbone-place water was raised by a pump, and supplied the portion of London in its immediate neighbourhood. Mr. Cavendish found that when boiled, it deposited a quantity of earthy matter, consisting chiefly of lime, but containing also a little magnesia. This he showed was held in solution by fixed air; and he proved experimentally, that when an excess of this gas is present, it has the property of holding lime and magnesia in solution.[187] Besides these earthy carbonates, the water was found to contain a little ammonia, some sulphate of lime, and some common salt. Mr. Cavendish examined, likewise, some other pump-water in London, and showed that it contained lime, held in solution by carbonic acid. [187] The salts held in solution are in the state of bicarbonates of lime and magnesia. Boiling drives off half the carbonic acid, and the simple carbonates being insoluble are precipitated.