In considering the effects of temperature upon the aneroid, they are found to be somewhat complex. There is the effect of expansion and contraction of the various metals of which the mechanism is composed; and there is the effect on the elasticity of the small portion of air in the box. An increase of temperature produces greater, a diminution less elasticity in this air. The compensation for effects of temperature is adjusted by the process of "trial and error," and only a few makers do it well. It is very often a mere sham. Admiral FitzRoy writes, in his _Barometer Manual_, "The known expansion and contraction of metals under varying temperatures, caused doubts as to the accuracy of the aneroid under such changes; but they were partly removed by introducing into the vacuum box a small portion of gas, as a compensation for the effects of heat or cold. The gas in the box, changing its bulk on a change of temperature, was intended to compensate for the effect on the metals of which the aneroid is made. Besides which, a further and more reliable compensation has lately been effected by a combination of brass and steel bars." "Aneroid barometers, if often compared with good mercurial columns, are similar in their indications, and valuable; but it must be remembered that they are not independent instruments, that they are set originally by a barometer, require adjustment occasionally, and may deteriorate in time, though slowly." "The aneroid is quick in showing the variation of atmospheric pressure; and to the navigator who knows the difficulty, at times, of using barometers, this instrument is a great boon, for it can be placed anywhere, quite out of harm's way, and is not affected by the ship's motion, although faithfully giving indication of increased or diminished pressure of air. In ascending or descending elevations, the hand of the aneroid may be seen to move (like the hand of a watch), showing the height above the level of the sea, or the difference of level between places of comparison." In the admiral's _Notes on Meteorology_, he says, "The aneroid is an excellent _weather glass_, if well made. Compensation for heat or cold has lately been introduced by efficient mechanism. In its _improved_ condition, when the cost may be about L5, it is fit for measuring heights as far as 5,000 feet with approximate accuracy; but even at the price of L3, as a _weather-glass_ only, it is exceedingly valuable, because it can be carried anywhere; and if now and then compared with a good barometer, it may be relied on sufficiently. I have had one in constant use for ten years, and it appears to be as good now as at first. For a ship of war (considering concussion by the fire of guns), for boats, or to put in a drawer, or on a table, I believe there is nothing better than it for use as a common weather-glass."