438 On the Passage of Air through Capillary Tubes. 





I. Pressure 315 millims. 



of Water. 





Temp. 

 Centigrade. 



Absolute 



temp. 



T. 



rj\2 



Time. 



rj\2 



*H. 



T 



T(T+94) 



0. 





t. 



t 



\t) 



r 



t 



-20 



253 



64009 



119 



538 



12 

 2 

 15 

 11 

 36 

 5 



2-42 



712 







273 



74529 



135 5 



550 



227 



713 



21 



294 



86436 



156-5 



552 



213 



705 



60 



333 



110889 



195 5 



567 



1-88 



703 



100 



373 



139129 



240-5 



578 



1-65 



696 



200 



473 



223729 



364-5 



614 



1-37 



712 



240 



513 



263169 



425 



619 



1-29 



713 





II. Pressure 1000 millims. 



of Water. 





20 



293 



85849 



110 



784-5 



120 

 100 

 7-9 

 8-8 

 116 

 7-9 

 3-8 

 9-4 

 9-6 



2-663 



1035 



40 



313 



97969 



123 



796-5 



2-545 



1036 



60 



333 



110889 



1375 



806-5 



2-422 



1034 



80 



353 



124609 



153 



814-4 



2-307 



1031 



100 



373 



139129 



169 



823 2 



2-207 



1031 



120 



393 



154449 



185 



834-8 



2124 



1035 



140 



413 



170569 



202-4 



842-7 



2040 



1034 



160 



433 



187489 



2215 



846-5 



1-955 



1030 



180 



453 



205209 



23975 



855-9 



1-889 



1034 



200 



473 



223729 



258-5 



865-5 



1-833 



1035 





1033-5 



These Tables show — 



(1) That the time varies approximately as the square of the 

 absolute temperature. This appears from the approximate 

 equality of the numbers in column 5. 



(2) That the variation of the time deviates from the law of 

 squares by a term approximately proportional to the tempera- 

 ture. This is shown by the average equality of the differences 

 in column 6, 



(3) That the formula connecting the time and the tempera- 

 ture is very nearly 



t=dT(T + /3), 



where fi seems to depend on the pressure pi —p 2 - 



From the above series of experiments, however, the exact 

 connexion between @ and p x —p 2 is not obvious ; possibly it 

 may be of the form 



*=*t(t+ ? g v 



The next series of experiments had in view the more exact 

 determination of the relation between pi —p 2 and t. 



To obtain these results, it was necessary to determine the 



• ■ 



