VISCOSITY OF WATER BY THE EFFLUX METHOD. 133 
Wagner, 1883.—Wagner’s measurements of efflux were all with 
one tube, four experiments being made at every 5° from 15° to 
50°. His apparatus was similar to Wiedemann’s' and Sprung’s, I 
assume, as I did in reducing Slotte’s work, that Wagner’s dimen- 
sions are evaluated for the one temperature, though definite 
information on the point is not given. He used Hagenbach’s 
correction, In regard to the pressures, I conclude that H indicates 
grammes per square centimetre. 
Reynolds, 1883.—Two of the tubes used by Reynolds were 
leaden pipes whose radii were 0°3075 and 0°635 cm. In these 
the fall in pressure between two points 152-4 cm, apart was 
measured by the vertical height of columns of water sustained 
at those points, i.e. by the ‘hydraulic gradient’ of the pipe. 
Though not exactly equivalent to the fall per unit of length 
expressed in grammes per square centimetre, the agreement is 
sufficiently approximate in relation to the order of precision in 
the data, to render the application of a correction unnecessary, 
In my reductions I have assumed the acceleration of gravity to 
be 981: cm. The temperatures do not appear to have been 
measured with great precision so that the deduced value of 7 must 
be considered as merely approximate. It is worthy of remark 
however that it does not differ more than 1/27 from the value 
obtained from a tube (M, Poiseuille’s) the area of whose section 
was only about 1/800000 of the larger of these tubes. 
Rénigen, 1884.—The object of Réntgen’s experiments was to 
“Scertain the effect of pressure on the viscosity. He measured 
the times of the efflux under pressures varying from 19 to 29 
atmospheres, as well as those under pressures of 1 atmosphere, 
= temperatures being approximately 6° and 11°C. There is 
Cae information to determine the absolute or relative 
Sat these tem peratures, nevertheless the effect of variations 
Pressure can be fairly well ascertained. The change in the 
ite with such variations is so slight that the correction for 
ip 
_  *°88. Annal. Bd. 99, p. 221, 1856. 
