CALIBRATION OF ELECTRICAL PYROMETERS. 
119 
time series of boiling-point experiments made in the glass boiling tubes. 
The data of Table 16 are again available with a few supplementary 
results to be inserted below; The variation of temperature during 
about an hour's uninterrupted boiling in sulphur in case of No. 17 is 
within £ per cent, of e 20 ; in case of No. 18 about £ per cent, of e 20 ', in 
case of No. 22 within J per cent, of e 20 for almost two hours of boiling, 
etc. The constancy is satisfactory, and since the variations contain 
incidental errors it is probable that the boiling point itself is reached 
to within two or three degrees. To test this supposition I reversed the 
procedure, and by means of the given constants measured the boiling 
point of sulphur with the following results : 
Table 21. — Constancy of temperature in boiling tubes. Sulphur. 
No. 
t. 
e. 
C20. 
, 
Time. 
° 0. 
Micro- 
volts. 
Micro- 
volts. 
o(7. 
h. m. 
17 
20.4 
4005 
4008 
445 
11 15 
20.4 
4005 
4008 
445 
23 
18 
20.9 
3974 
3981 
445 
11 25 
21.2 
3967 
3976 
445 
32 
22 
21.5 
4000 
4012 
416 
11 37 
21.7 
4000 
4013 
446 
40 
These results corroborate the statements made. All these data are 
obtained with older forms of apparatus. I believe that if I were to 
repeat them with the newer forms (Figs. 14, 15) the limits of error would 
be much decreased and the experiment materially gain in certainty. 
Remarks of the same nature as the above apply to mercury, for which 
the conditions of constant temperature are much more favorable. A 
long time series was made with element No. 22, where ebullition was 
purposely prolonged nearly five hours. The total variation of boiling- 
point is here about £ per cent, of e 20 . This variation, too, is incidental, 
and is largely due to the fact that here, as in the case of sulphur, a 
gentle current of carbonic acid circulated through the tube. In later 
experiments I observed an almost absolute constancy; but it is not 
necessary to multiply these data here. 
What has been said of sulphur and mercury is, of course, eminently 
true of lower boiling points, like aniline and water, as Table 16 aptly 
shows. In case of substances which boil at lower temperatures still, 
like alcohol and ether, special devices for condensing the vapor in the 
lateral tube, which projects above the boiling tube, Fig. 9, must be re- 
sorted to. 
Available substances of fixed boiling points. — With these results in 
hand I thought it advisable to test the availability of a few organic 
and other substances of fixed boiling points. Usually such substances, 
even liquids like aniline, char or become so heavily gummed during 
(773) 
