The term “pyrometer”—formerly applied to instruments designed to measure the expansion of solids—is now used to describe any device for determining temperatures beyond the upper limit of a mercury thermometer. This limit, in the common form, is the boiling point of mercury: 357° C. or 672° F. By leaving the bore of the tube full of nitrogen or carbon dioxide prior to sealing, the pressure exerted by the enclosed gas when the mercury expands prevents boiling; and with a strong bulb of hard glass the readings may be extended to 550° C. or 1020° F. Above this temperature the hardest glass is distorted by the high internal pressure, but, by substituting silica for glass, readings as high as 700° C. or 1290° F. may be secured. Whilst such thermometers are useful in laboratory processes they are too fragile for workshop use; and if made of the length necessary in many cases in which the temperature of furnaces is sought, the cost would be as great as that of more durable and convenient appliances. No other instrument, however, is so simple to read as the thermometer; and for this reason it is used whenever the conditions are favourable. The latest proposal in this direction is due to Northrup, who has constructed a thermometer containing tin enclosed in a graphite envelope, which is capable of reading up to 1500° C. or higher. This instrument is described on page 216. The origin and development of the science of pyrometry furnish a notable example of the value of the application of scientific principles to industry. Sir Isaac Newton was the first to attempt to measure the temperature of a fire by observing the time taken to cool by a bar of iron withdrawn from the fire; but, although Newton’s results were published in 1701, it was not until 1782 that a practical instrument for measuring high temperatures was designed. In that year Josiah Wedgwood, the famous potter, introduced an instrument based on the progressive contraction undergone by clay when baked at increasing temperatures, which he used in controlling his furnaces, finding it much more reliable than the eye of the most experienced workman. This apparatus (described on page 211) remained without a serious rival for forty years, and its use has not yet been entirely abandoned. The next step in advance was the introduction of the expansion pyrometer by John Daniell in 1822. The elongation of a platinum rod, encased in plumbago, was made to operate a magnifying device, which moved a pointer over a scale divided so as to read temperatures directly. Although inaccurate as compared with modern instruments, this pyrometer was the first to give a continuous reading, and required no personal attention. The expansion pyrometer—with different expanding substances—is still used to a limited extent.