86 G. H. KNIBBS. 
only to a first approximation: a higher approximation require 
that it be increased about one-eighth, and it will be shown that : 
Boussinesq’s theory is experimentally vindicated. In the same 
year (1891) Brodmann! published the results of his researches 4 
with the sphere and cylinder methods—(/) and (e)—and showed | 
were obtained for the viscosity according to the size of the sphere 
or of the cylinder, confirming Elie’s? observation that values for | 
the viscosity deduced from experiments with concentric spheres 
by Kirchhoff’s formula, increased with the velocity of rotation. — 
In 1892 Cohen,? employing the efflux method, continued the — 
research in respect of the influence of pressure on the coéfiicient, 4 
for the temperatures 1°, 15°, and 23° C., and for pressures up t 
atmospheres. His results therefore add to the material for 
comparison. 
Many other contributors either to the theory or observation of 
the viscosity of liquids, or cognate matters, might be mentioned~ 
as for example, Hiibener 1873, Baumgartner ’74, Guerout 74 and 
76, Villari 76, Koch ’81, Grossmann 783, Graetz ’88, Wirtz 89, : 
Whetham ’90, Briickner 90 and ’91, and Rayleigh ’92—but the 
above sketch of the course of the investigation embraces its main 
historical outlines and subsequent references will be made only 
such of the above as especially bear upon the particular cme 
of any question under review. 4 
Ii.—Tue Turory or Erryvux. 
fluid be established in a long cylindrical capillary tube, mee see 
1Un ntersuchungen iiber den Reibu ungscoéfiicienten von " Fliissighe! 5 
Le Gottingen, 1891,—Wied. Annal. Bd. 45, P- 1” : 
nee a 
4 Sitzber. Akad. Wiss. Wien. Bd. 40, p. 653, 1860. 
