1G4 
MR. H. L. CALLENDAR ON THE PRACTICAL 
But the comparisons of different platinum wires show that this formula cannot be 
absolutely accurate.] 
It is very convenient for practical purposes that the platinum thermometer agrees so 
closely with the absolute thermo-dynamic scale ; but, of course, such agreement does 
not much affect its value as a standard. We at present require a practical standard 
to which all measurements may be directly referred ; the question of its exact relation 
to the absolute thermo-dynamic scale may be matter for subsequent investigation, and 
can be best solved when we have arrived at a more accurate and extended knowledge 
O 
of the phenomena of heat, of the dependence of radiation, conduction, &c., on tempe¬ 
rature measured by a scale, which may be to some extent arbitrary, but which will 
have the incalculable advantage of uniformity, so that the results of different experi¬ 
mentalists will be accurately comparable. To show how imperfectly the air thermo¬ 
meter fulfils this condition, we may refer to the article “ Pyrometer ” in the ‘ Encyclo¬ 
paedia Britannica,’ 9th edition, 1885. 
IV. In comparing the platinum resistance thermometer with other instruments, it 
will be seen that it is essentially practical. The steam pressure thermometers 
advocated by Sir William Thomson in his article on Heat, in the ‘Encyclopaedia 
Britannica,’ are doubtless theoretically more interesting, and could be made of great 
utility in attaining known constant temperatures when the relation between steam 
pressure and temperature had once been determined for each case. But the most 
accurate way of comparing them with each other, and with the standard air ther¬ 
mometer, is by the intermediation of the platinum wire. Other instruments for 
measuring temperature have either a very small range or lay no claim to accuracy; 
for this reason it is obviously unnecessary to discuss the mercury thermometer or the 
thermopile, especially as their defects are so well known, and their applicability so 
limited, although each possesses distinct advantages for certain purposes. 
Platinum wire is well fitted for use as a temperature standard, it is still more 
admirably fitted for practical use in any experiments where accuracy is required. This 
arises chiefly from its adaptability. A length of wire is cut off and disposed in any 
way which may happen to be most convenient for the desired purpose. The observa¬ 
tions are simple and rapid, the corrections small and easily applied, the sources of 
error less than with any other kind of thermometer and more easily avoided. For 
radiation experiments it may be made more sensitive than the thermopile (cf. Langley’s 
Bolometer) ; owing to its great sensitiveness, it may be used to measure rapidly varying 
temperature without inaccuracy ; this is often desirable, as in calorimetry by method 
of cooling. There is, in fact, hardly any experimental investigation in which the 
measurement of temperature is necessary, which may not be more accurately and 
simply effected by means of a suitable platinum wire thermometer. 
[The results of the investigation considered as a research on the resistance- 
variation of metals tend to prove the existence of minute irregularities which 
