VISCOSITY OF WATER BY THE EFFLUX METHOD. 129 
taken into consideration.! Hagen’s values are so nearly equal to 
those of later determinations that corrections were unnecessary. 
The viscosity was calculated from the formula 
ghop Rk? 
ected ee 
The variations of the dimensions with temperature have not been 
taken into account, so that they are uncertain to the extent of 
the correction + ¢ see section 18 hereinbefore : this however will 
not materially affect the result, and the data are certainly not of 
sufficient precision to make the matter worth consideration. The 
results have been reduced from the temperatures quoted to 0° C. 
by formula (38). The value of g for Kénigsberg, where these 
measurements were made, has been assumed to be 981°44. 
Hagenbach, 1860.—Hagenbach reduced his own and other 
experiments by the application of a formula, which, for C.G.S. 
units and an absolute value of », may be written 
a ea aces ets RR ( 
1 FOL (40) 
This was theoretically deduced and is identical with (4) if k =m/8, 
hence & should be 0:140—see § 8. Hagenbach however used 
y 24°, 4¢.0-0992, instead. The lengths of his tubes were measured 
directly, and their radii computed from the contained volumes of 
mereary.! The acceleration of gravity is not given, but I assume 
| Wt to be about 981-0 centimetres. The value for n at 0° C, given 
; yg is the mean of the results for four tubes of different 
ii, 
Graham, 1861 -—Graham measured the efflux times with two 
—D and E—at frequent intervals of temperature, never 
ates than 5°, from 0° to 70° C., recording generally to whole 
amg a half second however being occasionally shewn. From 
83 values he gives for (), Z and R, I have reduced his observed 
‘eas formula (5), casting the results to 0°1 second. The two 
en ee anything like identical values for », the results 
: Archiv. fur Anat. u. Phys. 1861, p. 315. ie we 
Pogg. Annal. Bd. 109, p.398. 8 Ibid. p. 400. 
TJuly 3, 1995, 
