Diffusion of Liquids. 



323 



below. But perhaps it would be well to explain first how 

 this changing the velocity of the H 2 can be readily accom- 

 plished. Into the caoutchouc stopper of the lower tubulure of 

 a Mariotte's bottle was fitted a tube bent twice in planes at 

 right angles to each other, and having the outward end drawn 

 to a fine point. Now, turning the tube in the stopper has the 

 effect of elevating or lowering the orifice — that is, of diminish- 

 ing or increasing the distance between this and the lower end 

 of the vertical air-tube in the flask. Upon this distance de- 

 pends the velocity of efflux, that being proportional to its 

 square root; and hence to attain the desired velocity it is simply 

 necessary to move the orifice up or down, as the case may be. 

 By this means it is easy to obtain a velocity varying not more 

 than a cubic centimetre per hour for a week at a time. 



n* KC1 solution. 



Duration of experiment 



24 hours. 



H^ . Cub. cent. 



2 in 



cub. cent. 



per hour. 



62 



^ A g' N0 3 



58 

 57 

 47 

 37 

 30 



per hour. 

 1-69 

 1-60 

 1-54 

 1-69 

 1-69 

 1-62 



2wNaCl solution. 

 Duration of experiment 



48 hours. 



H 2 Oin 



cub. cent, 

 per hour. 



28 . 



30 . 



30 . 



64 . 



65 . 

 6Q . 



. 70 . 



Cub. cent. 



— AeNO a 



40 3 



per hour. 



. 2-74 



. 2-80 



. 2-73 



. 2-62 



. 2-73 



. 2-71 



. 2-64 



NaCl solution, nearly 2 n, 



Duration of experiment 



26 hours. 



Cub. cent. 

 H 2 O per hour. 



50 

 49 

 48 

 45 

 44 



Cub. cent. 



per hour. 

 . 2-45 

 . 2-41 

 . 2-38 

 . 2-47 

 . 2-50 



2n NaCl solution. 

 Duration of experiment 



5 hours. 



Cub. cent. 

 H 2 O per hour. 



80 



100 



Cub. cent. 



ro A « NO » 



per hour. 

 1-84 

 1-70 

 1-20 



From the first four of the above Tables it is seen that, for 

 velocities varying between 30 and 70 cubic centims. per hour, 

 the movement of the water exerts no disturbing influence. In 

 the fifth experiment the velocity of the water had been constant 



* n signifies <■ normal," i, e. 1 eq. in grams in a litre. 



