322 



NA TURE 



[July 31, i! 



As the air in a in this case has expanded isothermically , 

 the mechanic work is represented by the equation 



x = p„v„ log nat — , 



where p v are the initial pressure and volume of the air 

 in A, and v is its volume after the expansion. If we 

 suppose that a beam of sun rays during a minute has 

 fallen upon the blackened surface of a through a hole 

 of a square centimetre, and that the experimenter has 

 neutralised the tendency of the air to increase in tempera- 

 ture by careful manipulation of the stopcock c, which 

 allows the air to expand its volume as it displaces the 

 escaping mercury — suppose that in this way the cooling 

 effect of the expansion has neutralised the tendency of the 

 air to augment its temperature under the influence of the 

 heat radiation, so that the index of the thermometer has 

 been kept constantly at its initial point, then the entire 

 amount of solar heat imparted to the glass reservoir A is 

 converted into mechanic work by means of the isothermic 

 dilatation of the air, and the value of x found by the 

 above equation represents what is called by modern 

 scientists " the solar constant." ' 



It is obvious that the chief difficulty lies with the thermo- 

 meter. The indications of this part of the instrument must 

 be extremely sensitive (up to some thousandths of a degree 

 Centigrade) and instantaneous, in order that the experi- 

 menter maj- be able to regulate the expansion so that a real 

 isothermic dilatation takes place. I judged that only two 

 kinds of thermometers could suit the purpose, and tried 

 first a differential glass thermometer. This is in fact very 

 sensitive, but as the pressure in A diminishes during the 

 experiment, the bulb a expands somewhat, and this has 

 a disturbing influence upon the index. I next inserted in 

 A a network of very thin thermo-electric elements (combina- 

 tions of iron and German silver), and observed the alter- 

 ations of temperature of the air in A by means of a mirror 

 galvanometer. As I found the ordinary system of magnets 

 in galvanometers far too heavy for the instantaneous 

 deflections here required, I constructed a new kind of 

 galvanometer, whereof I give a schematic view in Fig. 2, 

 because I think that it may really do some good work 

 in other cases, as it proved to be extremely sensitive. 

 The dotted lines represent a system of two concentric 

 (annular) magnets made of steel springs (from watches), 

 each magnetised to saturation between the poles of a 

 powerful Plucker electro-magnet. They are combined in 

 the astatic manner, but the dimensions of the material 

 are chosen so that the inner magnet has just sufficient 

 force to keep the whole system in the magnetic meridian. 

 The figure shows the position of the insulated copper 

 wires relatively to the magnets. M is a mirror of very 

 thin silvered glass ; c c is a massive copper ring. I 

 tested the sensibility of the instrument by adiabatic ex- 

 pansion of the air in A. This was effected by opening 

 the stopcock c for a moment. The slightest dilatation of 

 the air in this manner immediately showed its cooling 

 effect by a deflection of the scale in the mirror, but as the 

 deflection soon brought the magnets out of the electric 

 field, the amplitude of the oscillation was, as 1 had calcu- 

 lated, not great. However, as the oscillations did not 

 cease instantaneously, I found the method impracticable 

 for continuous observation. I then abandoned the project 

 of regulating the isothermic expansion by means of a ther- 

 mometer altogether. - 



The next arrangement, which succeeded better, was 

 that shown in Fig. 3. Here the co-ordinates of the iso- 

 thermic curve are traced out beforehand on the rotating 

 cylinder F. As the mercury in B flows into D, it lifts 

 a float, which, by a combination of wire and blocks, 



1 Uncorrected. 



z _ I think the tiling will be very difiicuit to realise in this way. If another 

 indicator could be substituted for the galvanometer, for example, the Lippmann 

 capillary electrometer in the ingenious form devised by Chr. Loven, the ex- 

 periment would be very easy. But, unhappily, this 

 to thermo-electr' 



makes the cylinder rotate at a rate which is proportional 

 to the expansion of the volume of the air in A. Thus the 

 horizontal co-ordinate (v) of the isothermic curve is 

 represented. The vertical co-ordinate of the pressure of 

 the air in a (^) is represented by the height of the liquid 

 in the open branch of the manometer. The operator only 

 has the task to regulate the outflow of the mercury from 

 b to d by means of the stopcock c, so that the level of 

 the fluid in the manometer closely follows the isothermic 

 line drawn upon the paper envelope of the rotating 

 cylinder. This is not difficult after a little experimenting. 

 Whenever the level in the manometer shows a tend- 

 ency to rise above the isothermic line, there is a surplus 

 of heat in A waiting for transformation into work, which 

 can be effected by accelerating the outflow of the mercury 

 through C. The area contained between the initial and 

 final ordinates [p and p, represented by the positions of 

 the column of liquid in the manometer-tube relatively to 

 the cylinder at the commencement and the close of the 

 experiment] represents the value of the integral — 



-/* 



or the amount of mechanic work equivalent to the trans- 

 formed caloric energy. Thereby this method affords an 

 elegant manner of showing the actual transformation of all 

 kinds of heat into work to an auditory. In order to obtain 

 indications on a grand scale I always used H 2 S0 4 tinted 

 blue with indigo in the manometer. The rotating cylin- 

 der is about 2 m. high, and a quantity of heat of not 

 more than 876 gramme-calories, imparted by radiation or 

 otherwise to the air in A, makes the cylinder rotate 

 360 ", and the level of the liquid in the manometer sink 

 1 '84 m. 1 The volume of A was 400 c.c, and the initial 

 pressure equal to 1000 mm. (of mercury). But for scien- 

 tific measurements I cannot recommend this method. The 

 sulphuric acid adheres to the glass tube, and does not take 

 up its definitive level at once, the dimensions of the 

 apparatus become inconveniently large, the co-ordinate 

 p cannot be traced out on the paper of the cylinder 

 directly from the isothermic equation, 

 pv = RT w 



but must be recalculated with the aid of some corrections 

 arising from the influence of the atmospheric pressure 

 upon the columns of liquid in the manometer and in D, 

 too complicated to be mentioned here. 



Fig. 4 shows a kind of calorimeter which realises the 

 condition of isothermic dilatation of the air in the most 

 simple manner and still is capable of the most accurate 

 measurements. 



A and A, are very thin glass vessels fabricated of equal 

 shape and size by Franz Midler in Bonn. Both contain 

 dry air over mercury, which stands at equal height in B 

 and Bj, If the graduated glass tube D, which communi- 

 cates witli A and A 2 through a caoutchouc tube, is raised 

 or lowered by means of the arrangement shown in the 

 figure, the level of the mercury rises or sinks equally in 

 B and B,, and the air in a and a, is compressed or dilated 

 equally, provided that the temperature is kept constant in 

 both. This condition is realised in A, by the surrounding 

 large mass of water, which imparts to the air and mercury 

 in A, and B, its own constant temperature. The air in a, 

 therefore always expands or contracts isothermically. If 

 its initial volume and pressure are denoted by v and p„, 

 the law of Mariotte, 



Vp =7'„Aj; 

 will always regulate the expansion of the air in a,. It is 

 easy to see that this will also be the case in a, if the experi- 



for the pres; 



cally. This height is 



vhat reduced by the i 



