130 G. H, KNIBBS. 
for 10° ©. being, for tube D -0112, and for tube Z -0081, instead 
of ‘0131. Isuspect that the dimensions are in error, which if — 
true, prejudices the correction of the times—the reduction of T' to 
7’,—a correction of considerable magnitude in these experiments 
Despite this, they are valuable as to some extent confirming— — 
contrary to Rosencranz’s measurements—the propriety of extt 
polating beyond its major limit, the curve expressing the results : 
from 0° to 45°C. Both bulb and tube were immersed in the 
heating bath, so that no dimensional temperature corrections 
were required. 
Rellstab, 1868.—According to the specific viscosities quoted by 
Wagner,! Rellstab’s experiments give for the relative fluidities for : 
each 5° from 0° to 50° the following values, that at 0° being 1000, 
1000, 1172, 1361, 1587, 1802, 2053, 2222, 2500, 2688, 2899, 
3205. These agree well with Poiseuille’s values, but in the absen® 
of the experimental data I have been unable to apply the revised 
correction for fall in pressure at the tube entrance. | 
Sprung, 1876.—Sprung’s observations have been reduced 0B 
the assumption that the dimensions of the tube and bulb are gi¥@ — 
pressure in grammes per square centimetre. Sprung used 
Hagenbach’s correction to the pressure.2 The reservoir and 
capillary being both immersed during the experiments, dimens! 
corrections were not required.® 
Rosencranz, 1877.—In Rosencranz’s apparatus the reservoir | 
supply, a bulb by means of which the volume of efflux bigs 
capillary. This complicates the reduction. Let @ denote 
volume of the reservoir at the temperature for which the dimen 
sions of the tube are given, and p the density of water at oe 
temperature, then the effluent volume Q’ at 7 degrees abo 
temperature, the density being then p’, will be 
1 Wied., Annal. Bd. 18, p. 263. 
2 Pogg. Annal. Bd. 159, p. 6. 3 Ibid, p. 5, 6. 
