Transmission of Gases through Water, 369 



The following experiment was also made with distilled 

 water, wide bore, 3 mm., inlet and exit tabes being employed. 

 The inlet-tube dipped 25 mm. below the surface of the 

 water : — 



No. 12. — A tube 1980 mm. long and 21 mm. bore was filled 

 with nitrogen-free distilled water, and the air-current con- 

 tinued for 12 hours. Temperature of room 15°*5 C, of 

 jacket 12°-9 C. 



Depth of Layer. 



to 200 10'56 



650 to 800 3-9 



1500 to 1600 2-68 



Before aeration 1*51 



The question arises, what is the active cause of the 

 streaming ? The author does not desire at this stage of his 

 investigation to definitely advance any explanation of the 

 facts observed. One, however, which is suggested by the ex- 

 periments that have been described, may be pointed out. It 

 is that the streaming is possibly caused by minute dust 

 particles, or by other centres of condensation possibly of an 

 electrical nature, carried by the air-current and being taken 

 up by the water together with the gaseous constituents of the 

 air, and that these bring about in some way a sufficient 

 density of the dissolved or liquefied gases to render it possible 

 for them to be drawn gravitationally downwards through the 

 Avater. 



The fact that the streaming is more marked in sea than 

 in distilled water may be explained by assuming that the 

 sea-water, from the salts which it holds in solution, is able to 

 make more use of the centres of condensation introduced by 

 the air which is bubbled through it. When sea-water and 

 distilled water are separately shaken up with the air, the 

 difference of behaviour of the two waters is, as is well known, 

 very marked, it being easy to obtain a formation of small 

 bubbles of air in sea-water to an extent quite impossible in 

 distilled water. 



There can be no question, from the experiments recorded, 

 that if the air be completely washed, the surface of sea-water 

 exposed to it will become saturated with nitrogen, but no 

 streaming effects will take place. 



It is equally evident from these experiments, that whatever 

 may subsequently be discovered to be the true cause of the 

 streaming, its effect in large volumes of sea- or river-water 

 under natural condition- must be of great importance, and of 

 such dimensions that the effect of ordinary diffusion may, in 

 comparison, be entirely neglected. 



PhU. Mag. S. 6. Vol. 9. No. 51. March 1905. 2 B 



