Adeney and BiiCKVAi— Solution of Nitrogen and Oxygen. 149 



In order to reduce these experiments to a basis which will allow com- 

 parison with the previous experiments, the values of /have been calculated 

 in the usual way, assuming that the process of solution in this case is the 

 same as before. These values of /are given in the last column of Table III, 

 and on comparing them with the values already deduced for tap-water and 

 sea-water some interesting relations become evident. 



It will be seen that the value of / for the tube through which undried 

 air was drawn is approximately the same as that derived for tap-water from 

 previous experiments, but the value for the tube with the dried air is very 

 much greater, and that for the tube with the air saturated with aqueous 

 vapour is very much less than the mean value. This shows that when the 

 air above the water is very dry the absorption of atmospheric gases takes 

 place very much more rapidly than usual. 



On the other hand, wlien the air is saturated, or nearly saturated, with 

 aqueous vapour the rate of solution is very greatly retarded. This would 

 seem to indicate that the process by which the dissolved gas is carried down 

 into the body of the liquid is affected by evaporation from the surface of the 

 liquid, since the evaporation will be at a maximum when the air is dry, and 

 a minimum when it is moist. 



In considering, therefore, this question of the mechanism of solution, there 

 are two factors which, whatever others may affect it, are obviously of primary 

 importance, namely : (1) concentration of dissolved salts in the surface- 

 layers due to evaporation ; (2) cooling of the surface-layers produced by 

 evaporation. 



Both these factors are of such a nature as to result in an increase of 

 density of the surface-layers, which would tend to set up vertical currenl,s in 

 the body of the water, hi the case of waters containing large quantities of 

 dissolved salts, the concentration of these salts by evaporation must play a 

 very important part in the process of solution of air. Jn this connexion it 

 is important to notice that the rate of solution of air by quiescent bodies of 

 sea-water may be much more rapid than that by similar bodies of fresh 

 water, as shown by the relative magnitudes of the values of / in each case. 



In the case of water containing only small quantities of salt in solution 

 the effect of concentration would not be so great, but would probably be of 

 somewhat the same magnitude as the effect produced by the cooling of the 

 surface layers. This cooling will of course, result in an increased density 

 of the surface layers, provided the temperature is not below 4"C., and 

 consequently it will in most cases operate in such a way as to hasten the 

 process of solution. 



These two factors, tending as they do to set up a slow circulation from the 



