RECENT PROGRESS IN PHYSICAL SCIENCE. 115 
In regard to the purity of the wire, it may be remarked 
that, although the temperature coefficient is appreciably 
diminished by slight impurities, the effect on the indicated 
temperatures is quite insignificant unless the amount of the 
impurities is considerable. 
Special attention is invited to resolution No. 4 as proposed, 
which specifies that the normal boiling point of sulphur on 
the gas-tliermometer scale be taken at 444°.53 C. This is 
the value found by Messrs. Callendar and Griffiths by the 
use of a constant-pressure air thermometer.* Harker and 
Chapuis, however, in the course of an elaborate comparison 
of the platinum thermometer with the constant-volume ni¬ 
trogen thermometer at the International Bureau of Weights 
and Measures, have arrived at the value 445°.27.t The 
discrepancy of about three-fourths of a degree, centigrade, 
between these two results is not thus far explained, and is 
only partly due to well-known differences between the two 
types of gas thermometers employed. 
Enough has been said to show the place platinum ther¬ 
mometers are sure to take in future scientific research. The 
extreme precision and delicacy of electrical methods insures 
great accuracy and exceeding sensitiveness, and that, too, in 
an instrument of very small mass and heat capacity. In 
many details the instrument is peculiarly adapted to a wide 
range of research where accuracy is of importance. 
Before concluding this very imperfect account of the re¬ 
cent progress in instrumental physics, I desire to ask your 
attention to the tendency of the times as regards steam 
motors. The steam-engine can hardly be classed as a phys¬ 
ical instrument; nevertheless the phenomena of heat involved 
in its action bring it within the broad domain of physics, 
and will, I hope, justify noticing here the direction in which 
progress is now being made. 
For over two hundred years mechanical engineers, with 
scarcely any notable exception, have expended their energies 
*Phil. Trans. Roy. Soc., A, 1891. 
fPhil. Trans. Roy. Soc., A, 194, 1900. 
