302 MEASUREMENT OF HIGH TEMPERATURES. [bull. 54. 
ration method were to be rejected, it seems to me that very little ex-, 
perimental elucidation of the thermal variation of gaseous viscocity 
will be obtainable. 1889.] 
Again, the quantity actually measured is not rj but — %- , where C is 
the coefficient of slip (Gleitungs coefficient) ; % being inversely pro- 
portional to pressure 1 , or according to Meyer 2 and others, a direct ex-l 
pression for mean free path. Hence it must have a smaller value in the 
case of platinum tubes than in the case of tubes of glass, because of the 
tendency of gas to condense on metallic (platinum) surfaces. What- 
ever discrepancy is introduced by C will therefore be more serious in 
case of glass than in case of platinum. From this point of view results 
with metallic capillary tubes have greater claims to correctness. Fur- 
ther remarks on this quantity £ I have given elsewhere. 3 
THE NEW METHOD OF PYROMETRY. 
PRACTICAL REMARKS. 
Appurtenances. — Having endeavored to investigate the theory of the 
transpiration thermometer it will be expedient to close this chapter 
with a few practical remarks. Equation 12, on page 281, shows that 
thermal data can be measured in terms of time, t, or of volume, V 0j or 
again in terms of the rate at which volume increases, - .° -"; or even in 
P— v 
terms of the pressure factor, -• Any of these quantities varies as 
the f power of absolute temperature, and hence the sensitiveness of 
the method. As regards expedition, however, the differential methods 
are possibly more serviceable; and either the volume method which I 
employ on page 293 or Mr. Holman's method used either with compres- 
sion or exhaustion is available. Again, if a suitable jet-pump is not at 
hand and if the mercury apparatus, page 243, Fig. 43, presents too 
much practical inconvienence, a bell-jar arrangement, like the gasome- 
ter, Fig. 35a, page 179, in which the light bell may be weighted either 
positively or negatively, suggests itself. Such an apparatus may of 
course be made small and refilled for each measurement. Differential 
apparatus obviate the use of chronometers and of pressure measure- 
ments, and solution and other errors are similarly eliminated. 
The transpiration pyrometer. — To fully utilize the accuracy of measure- 
ment of which the transpiration pyrometer is capable, it is necessary to 
give the instrument a different form from that used in the present pio- 
neering experiments. Inasmuch as it was my object to study effects in- 
i Kuudt u. Warburg : Pogg. Ann., vol. 155, 1875, pp. 337, 525. 
3 Meyer: Kinetische Theorie d. Gase, p. 152. 
Barus: Wied. Ann., vol. 36, 1889, p. 383. 
(956) 
