230 Johnston and Adams — High Pressures on the 



salt solutions and turpentine, but a decrease with water and 

 with the weaker solutions at the lower temperatures. The 

 variation in the case of water is not proportional to the pressure, 

 and is greater at the lower temperatures, but Cohen did not 

 observe a minimum viscosity. These results were confirmed 

 by Hauser,* who investigated the viscosity of water at pressures 

 up to 500 atm. and temperatures up to 100°. He found that 

 the effect reached a minimum at 32°, at which temperatures a 

 pressure of 400 atm. was without effect on the viscosity ; 

 above 32° pressure causes an increase in the viscosity of water. 



All the available evidence indicates, therefore, that the effect 

 of pressure is to increase the viscosity of liquids. The anoma- 

 lous behavior of water at low temperatures is in line with its 

 abnormal behavior in other directions 7 and is, in all probabil- 

 ity, altogether exceptional. 



Kontgenf also investigated the effect with marine glue,f a 

 substance which behaves as a very viscous liquid or as a brittle 

 solid according to the character of the forces to which it is 

 subjected. He made parallel experiments on the penetration 

 of a block of the material by a loaded rod (a) at 1 atm., (b) at 

 500 atm., and found that its viscosity increased under com- 

 pression, but regained its original value when the pressure was 

 removed. 



These comprise, so far as the writers are aware, all the investi- 

 gations dealing with this question ; they show that the effect of 

 uniform pressure is always to increase the viscosity (water and 

 certain dilute aqueous solutions, both at low temperatures, 

 excepted). 



The question of the viscosity of solids is as yet by no means 

 in a satisfactory state ; but, for various reasons, we cannot 

 discuss§ it here. Suffice it to say that we have at present no 

 accurate knowledge of solid viscosity | (the term even is ill- 



* Ann. Phys., v, 597, 1901. \ Ibid , xlv, 105, 1892. 



% Barus, using a transpiration method, measured the viscosity of marine 

 glue at ordinary pressure and temperature, and found it to be of the order 

 of 10 8 as compared with values of 10 -2 for water and about 10 for glycerine, 

 at ordinary temperature. (IT. S. Geological Survey. Bull. 73, 134, 1891.) 

 The brittleness of this substance, noticed when even moderately great forces 

 are suddenly applied, is correlated with its high viscosity. 



§ The reader desirous of further information will find the earlier work (up 

 to 1890) reviewed by Barus, in his bulletins on solid viscosity (No. 73 cited 

 above, and No. 94, 1892 ; the greater part of this work appeared also in this 

 Journal, and in Phil. Mag., 1888-1892). A recapitulation of the present-day 

 status of the question will be found in text-books of elasticity, e. g. " The 

 Mathematical Theory of Elasticity." either of Love (Cambridge University 

 Press, 2d edition, i905), or of Ibbetson (Macmillan) ; or in Winkelmann, 

 " Handbuch der Physik" 2 Aun. 1908. 



|| In this connection it may be observed that any flow of a solid is accom- 

 panied by a change in its molecular configuration and in its properties ; in 

 all such cases, therefore, we are not dealing with an absolutely definite sys- 

 tem. The same remark applies more or less to measurements of the tensile 

 strength and analogous "constants " of metals. 



