500

Fixed Points in the Thermometer.

lower point without solidifying than any other, except alcohol or ether. Its little capacity for heat, and its great conducting power, cause it to be very quickly affected by changes of temperature. Again, its expansion is singularly equable for an equal increase of heat through the important middle part of the scale, which includes all the common temperatures on earth, namely, from the freezing to the boiling heat of water. Owing to the non-adhesion of mercury to glass it may be used in very minute tubes, as capillary attraction does not interfere with its movements. It is easy to proportion the bulb and the tube to each other, so that a small difference of temperature shall cause the mercurial column in the tube to rise or fall very conspicuously.

Now when the important fact was ascertained that solid water or ice always melts at precisely the same temperature, and that liquid water boils always at the same temperature, it followed that, by placing such a thermometer as above described, first in melting ice, and then in boiling water, and marking upon the tube the two points at which the mercury stands, as represented here by F and B, two fixed or invariable points of reference are obtained, and the interval between them can be divided on the glass, or on a suitable scale attached to the glass, into any convenient number of parts to be called degrees. If follows farther, that by continuing the divisions to any extent both above and below the fixed points, a general scale of temperature would be obtained, with respect to which all thermometers made on the same principle would perfectly agree, although the size of the divisions on the tubes, would vary in the different instruments according to the comparative capacities of the bulb and tube.

713. The space between the two fixed points has been variously sub-divided, i.e., there has been no agreement as to what should be understood by a degree of heat. Hence we have three different scales-1st. The Centigrade, so named from its being divided into 100 spaces or degrees between freezing and boiling water. This was introduced by Celsius, a Swedish professor, in 1741. It is chiefly used in France and Northern Europe. 2nd. Réaumur's Thermometer, used in Spain. Réaumur, an eminent French philosopher, in 1731, divided the interval into eighty parts or degrees. This thermometer is but little employed, as the degrees are inconveniently large, and involve the use of fractions. 3rd. Fahrenheit, so named after its inventor, an ingenious experimental philosopher and Fellow of the Royal Society, born at Dantzic in 1686. In

The Three Different Scales.

501 this scale, which was brought out about 1714, the space between boiling and freezing water is divided into 180 parts or degrees. Fahrenheit's thermometer is generally used in Britain, Holland, North America, and by English-speaking nations, and has many advantages over the other two. It is generally employed in manufactories, and for pharmaceutical processes. Its use is also sanc. tioned for fiscal purposes by Acts of Parliament.

The relations of these thermometers are of some importance, and it will be perceived from the different divisions of the scale that 9° of Fahrenheit are equal to 4° of Réaumur or 5° of Centigrade. Therefore, according to the rule of three, multiplying by nine and dividing by five or four, or the reverse, and adding or subtracting the 32° of Fahrenheit when required, gives the equivalent degree. It also follows from this relation, that the sum of the divisions of Centigrade and Réaumur (100 + 80 180) corresponds exactly to the divisions or degrees of Fahrenheit's scale. If, therefore, the degrees of C. are added to those of R., and 32° is added for temperatures above freezing, or deducted for temperatures below freezing, we obtain the degrees of Fahrenheit. Thus 21°C. + 17°R. + 32°

=

=

32°

=

70° Fahrenheit, and 18°C. + 14°R. o, or zero of Fahrenheit. A real zero implies that point at which bodies lose all heat. From certain data it has been calculated to exist at about 460° below the zero of Fahrenheit (Ganot). This of course has not been verified by experiment.

In the Centigrade and Réaumur thermometers the freezing point of water is taken as zero, while Fahrenheit places his zero at 32 of his degrees below this point.* He made a mixture of snow and sal-ammoniac, and from this deduced his zero by the degree of cold produced.†

The zero thus taken by Fahrenheit is purely artificial, as with

* Experiments lately performed by Mr. Sorby show that the temperature at which water freezes may vary. In tubes of small diameter, water may be cooled to 23° without freezing, and in capillary tubes, it required to be cooled to 104 in order to pass to the state of ice. This is close to the zero of Fahrenheit, and more than 30° below the freezing point of water, assumed as zero in the Centigrade and Réaumur scales.

† Fahrenheit found by experiment that 11, 124 parts of mercury, raised from the degree of cold produced by a mixture of snow and sal-ammoniac, to the temperature of boiling water, increased in volume to 11,336 parts. The difference, 212, was taken by him to represent the entire range of the scale from his assumed zero to boiling water.

502

Defects of the Centigrade Scale.

our present knowledge every zero must be. Graham has compared the scale of temperature to a chain extended both upwards and downwards beyond our sight. We fix upon a particular link and count upwards and downwards from that link, and not from the beginning of the chain. Fahrenheit preferred small to large links, and he placed his fixed point or zero so far below melting ice that all ordinary temperatures in habitable climates, are at once indicated by figures without rendering it necessary to resort to the + and signs. In assuming melting ice as zero, the Centigrade and Réaumur thermometers require the constant introduction of these signs, leading to the risk of omissions and mistakes. Thus 14° F. stands for 18° below melting ice, but 14° C. may mean either 57° F. or 7° F. according to whether it is above or below melting ice. It is a great defect in these two instruments that, owing to this malposition of zero, there should be for common temperatures, a necessity for counting upwards and downwards. Further, the degrees are on much too large a scale for common use. Each degree of C. represents (108) nearly 2° of F., and each degree of R. 24° of F. Hence slight changes of temperature are indicated on Fahrenheit's scale by whole degrees, when on the two other scales, fractions must be resorted to. The great inconvenience of this will be rendered obvious by the following comparative table showing the range of temperature at Eastbourne for the last seven days of January, according to Dr. Allnutt's observations. Mr. Symons has quoted this table to show how ill-adapted the Centigrade scale is for meteorological observations.

Typographical errors are frequently made in the + and and thus cause great confusion.*

* The Centigrade thermometer has been described as the more scientific instrument, because it has a fixed and definite zero at melting ice. It may

Test of a good Thermometer.

503 714. The test of a good mercurial thermometer is that the tube should be of equal diameter throughout. A short column of mercury moved up and down should have the same exact length in all parts of the tube. No air should be left in it, or this will form a resistance to the free expansion of the mercury, and impair the accuracy of the instrument. In inverting a good thermometer, the mercurial column either fills the tube or breaks off at the bulb and falls to the end of the tube. This shows a perfect vacuum. If it does not fall completely, or breaks into several columns, the tube contains air. If it has been properly graduated,—when the bulb is placed in melting ice the mercury will fall to 32°, and remain fixed at this degree until all the ice is melted. Water may be cooled to 30°, and even lower, without freezing, but melting ice always represents a temperature of 32°. Out of eight thermometers the writer has found only two accurate in this respect. Many errors are made in recording temperature, owing to thermometers not having been thus tested and compared.

715. Although the direct expansion of any solid body by a moderate change of temperature, is so inconsiderable as to be with difficulty measured, M. Breguet, of Paris, ingeniously contrived, in 1823, a solid thermometer which makes it very evident. It consists of three thin layers of silver, gold, and platinum, the gold being in the centre. These are laminated into a delicate metallic ribbon which is twisted in the form of a spiral, and fixed at one end, the other end carrying a light copper needle, which can move round a horizontal dial graduated on the Centigrade scale. The different rates of expansion of the three metals by slight changes of temperature, are indicated by the movements of the

be observed, however, that it is hardly scientific to use mercury for the measuring liquid, and to take for zero the solidifying point of water, which renders it necessary to count upwards and downwards for the ordinary range of temperature. Had the freezing point of mercury been selected as zero, there would have been some claim to scientific consistency in the arrangement, and + and signs would have been thereby avoided. The freezing point of this metal is 72° F. or 40° C. below freezing water, for 40° × 1.8° 72°. A thermometer thus constructed in F. degrees would have the melting point of ice at 72°, and boiling water at 252°. The objection to such a thermometer, is that it would make the scale inconveniently long for general use.

=

504

Differential Thermometer.

needle, the spiral either contracting or opening accordingly, as a result of unequal expansion between the silver and the platinum.

716. Air, although not used in ordinary thermometers, is well adapted to the formation of a thermoscope. It has a great extent of dilatation from a small increase of heat; it quickly receives impressions, and its dilatation is equal for equal increments of heat at all temperatures :—but atmospheric pressure cannot be excluded without at the same time confining the air and affecting its expansion. Professor Leslie, however, devised for particular purposes an airthermometer, which he employed very usefully, calling it the differential thermometer. It consists of two bulbs, a and b (fig. 178), filled with air, and connected by a bent tube, d c, containing liquid—the instrument being hermetically sealed, so that the atmosphere cannot affect the air within. A greater heat in the bulb, b, than in a, as when b is touched by the warm hand or is exposed to the sun's rays, causes the liquid to descend in the tube, d, which has a graduated scale attached to it. We may observe that equal divisions or degrees marked on the scale of this thermometer, do not mark equal changes of temperature, for the increasing condensation and resistance of the air in the colder bulb, and the increasing height of the liquid column to be lifted, requires the force overcoming these to increase progressively.

Fig. 178.

717. Maximum and Minimum Thermometers.—It is often desirable, as in garden hot-houses, to know how low the temperature has fallen during the night when no observer was present, and how high it has risen in the day. Both ends are attained by the use of Rutherford's twin-thermometers with horizontal tubes. The mercury in the tube of one pushes forward, as it advances with heat, a small piece of steel wire, and leaves it as a marker unmoved when it recedes again. The spirit or alcohol in the other, when contracting with cold, draws after it, by the mutual attraction, a small thread of coloured glass as marker, and then flows past this without moving it, when again expanding by heat.*

* Mr. Negretti has constructed a thermometer to record temperature for any hour or series of hours on the principle of his deep-sea thermometer (see Art. 577, p. 391). The instrument is fixed on a clock, which can be set