502 Scientific Intelligence. 
8. Measurement of high Temperatures.—In a second paper L. 
Hotsorn and W. Wien discuss the use of the increased resist- 
ance of platinum, and the use of a thermal junction of platinum 
and platinum-rhodium (10 per cent rhodium). They enter into 
a discussion of the constancy of the indications of the thermal 
junction. The determination-of a number of boiling points of 
different metals by different methods was undertaken, and it was 
shown that the thermo element gave very constant indications, if 
it is protected from the action of carbon; and that the formation 
of platinum silicate at high temperatures in the protecting cylin- 
ders of porcelain or clay do not influence it to a marked degree. 
It was found with the method of the resistance of platinum, 
that the temperature-coefficient of the resistance was much more 
sensitive toward outside influences than the thermo element and 
that many precautions had to be taken to measure this coefficient. 
The formation of platinum silicate lowers not only the tempera- 
ture-coefficient but also increases the resistance of the platinum. 
By means of an oven of clay of peculiar construction a tempera- 
ture of 1570° was reached. The following table of melting points 
gives their results in comparison with other observers: 
Holborn 
Violle. Barus. Barus. and Wien. 
Silver Eo et 954° 986° 985° 971° 
GroldG. At uee 1045 1091 1093 1072 
Copper LBS op bh, 1054 1096 1097 1082 
Nickel ie 2 us Bot) 1476 1517 1484 
Palladium _.-. 1500 1585 1643 1587 
Platinum. 2220! Dy ar gs: 1757 1855 1780 
Violle used a calorimetric method and interpolated the specific 
heat of platinum from 1200° upward. . 
Barus used a thermo element of platinum iridium and placed 
the element not in the interior of the air chamber but outside. His 
measures extend to 1050°. The higher melting points are extra- 
polated. Holborn and Wien claim that very different temperatures 
are obtained in the inside and outside of the chamber. In order 
to obtain a better equilibrium of temperature, Barus causes the 
muffel of the oven to rotate. The constancy of the thermo ele- 
ment used by the authors of this paper was tested by the com- 
parison of many determinations of melting points. With differ- 
ent thermo elements the determination at 1000° did not vary more 
than --5°. The thermal function does not vary from year to year 
or from use or disuse. 
Keiser and Schmidt of Berlin make, under direction of the 
authors, a D’Arsonval galvanometer of which the pointer moves 
over two scales, on one of which can be read the difference of 
potentials in volts; on the other the degrees of temperature indi- 
cated by the thermo element. The paper concludes with a state- 
ment of the superiority under different conditions of the method 
of the thermo element over the method of the resistance of 
