baeus.] VISCOSITY OF GASES. 281 
THE NEW METHOD OF PYROMETRY. 
Methods of computation. — In view of these harmonious results, I am 
induced to place great reliance on the accuracy of the relation 
?f = r /0 (l + ad")i, 
as a law, which in the case of a diatomic perfect gas like air or hydrogen 
expresses the thermal variations succinctly. Subject to the validity of 
this law, the favorable character of my results therefore introduces a 
new method of high-temperature measurement ; for by applying the 
Poiseuille-Meyer equation to transpiration data, such measurements 
can be made absolutely, throughout a wider thermal range, and with 
much greater convenience and accuracy than is the case with any other 
known method, not excepting the air thermometer. Moreover, the ex- 
clusive dependence of thermal data on the coefficient of thermal ex- 
pansion can now be put to the test inasmuch as a series of analogous 
data may be investigated on the basis of the above relation, and the 
two series of data can be compared throughout the whole interval of 
temperature observed. 
Introducing t/ / =r/ (l+a0")imto Meyer's equation, and solving with 
respect to 0", the following form results : 
C" 
(l+ad")* __n 1+4 "W'P 2 — p 2 t" R Q "* 
(1+/J0")<~16 7/0 7tf — V I" 
m 
1+4 1 •* 
or more simply 
Here t/ is the zero-viscosity of the gas selected for temperature 
measurement. 6 denotes the temperature of the cold ends of the capil- 
lary platinum tube, and y is Mr. Holinan's coefficient for air, approxi- 
mately ^=0.00275. The remaining variables have the signification 
given on page 253. R Q " and R being respectively the radii of the hot 
part {I"), and cold parts (Z', V") of the capillary tube at 0° 0., are prac- 
tically equal in my apparatus. Hence, to reproduce the temperature, in 
Tables 86 to 89 from the measurements of viscosity made, the equation 
takes the simple form 
(1+y* #£-16.76. V o V V \* * J I" U+U.UUb4 ») . (W) 
OF 
. (»35) 
