"During the year 1844 some experiments were made by James Glaisher, Esq., F.R.S., on the temperature of the water of the Thames near Greenwich at the different seasons of the year; when that gentleman found that the indications of temperature were greatly affected by the pressure on the bulbs of the thermometers. At a depth of 25 feet this pressure would be nearly equal to the presence of three-fourths of an atmosphere. These observations demonstrate the importance of using in deep-sea soundings an instrument free from liability of disturbance from compression by the surrounding water, and have ultimately led to the construction of the thermometer now to be described. "The instrument is composed of solid metals of considerable specific gravity, viz. of brass and steel, the specific gravity of these metals being 8.39 and 7.81 respectively. They are therefore not liable to compression by the water, which under a pressure of 1,120 atmospheres, or at a depth of 5,000 fathoms in round numbers, acquires a density or specific gravity of 1.06. In the construction of this instrument, advantage has been taken of the well-known difference in the ratios of expansion and contraction by heat and cold of brass and steel, to form compound bars of thin bars of these metals riveted together; and which will be found to assume a slight curve in one direction when heat has expanded the brass more than the steel, and a slight one in the contrary direction when cold has contracted the brass more than the steel. "The indications of the instrument record the motions under changes of temperature of such compound bars; in which the proportion of brass, the more dilatable metal, is two-thirds, and of steel one-third. [Illustration: Fig. 70.]