Adeney and Becker — Solution of Nitrogen and Oxygen. 391 



tion took place in a much shorter time, and it was found possible to saturate 

 the water in about twenty-four hours. 



This modification in the, method of employing the tubes brought the 

 series of observations under better control, and more concordant results were 

 obtained; nevertheless the observations showed some irregularity, which, 

 though small, could not be regarded as negligible. Attempts were, therefore, 

 made to work out an entirely new method of experimenting, and to devise 

 some method by which the water could be uniformly mixed during aeration, 

 and still preserve an unbroken water-air surface. 



One of the simplest means of effecting this seemed at first to he to pass a 

 known volume of air, in a slow and uniform stream of bubbles, through a 

 given thickness of water, in an observed time, and to determine the amount 

 of air dissolved during that time. Experiments carried out in this way 

 yielded results which afforded a remarkably uniform curve ; but attempts to 

 gain an accurate estimate of the water-air surface exposed proved unsuccess- 

 ful, owing to the variation in the size and shape of the hubbies, as well as to 

 the irregularity of the paths they took in ascending the water. 



With a view to bringing the method under better control, efforts were 

 made to direct the course of the bubbles into a straight line, and it was 

 found that the conditions prevailing, when a large bubble is allowed to 

 ascend a narrow tube filled with water, were almost exactly those sought. 

 In this case, the surface tension keeps the water-air surface unbroken during 

 its passage up the tube, and since the bubble occupies nearly the whole 

 cross-section of the tube, the water is exposed to the air in an extremely 

 thin layer; while the upward movement of the bubble produces a very 

 perfect local mixing of the water. 



Since the bubble tends to assume a simple geometrical form, the calcula- 

 tion of the area exposed, from measurements, is rendered comparatively 

 simple, provided that the measurement of length is made when the bubble is 

 in motion. In practice this was achieved by taking an instantaneous photo- 

 graph of the bubble with a glass metre scale placed alongside of it, in the 

 same focal plane. 



Some difficulty was at first experienced in getting the edges of the 

 meniscus well defined ; but this difficulty was finally surmounted by colour- 

 ing the water slightly with eosin, and taking the photograph by transmitted 

 light from a powerful arc-lamp. 



After some trials, it was found that the simplest method of measuring 

 the amount of absorption of air, after each excursion of the bubble, was by 

 measuring the loss of pressure exerted by the bubble. 



The form of apparatus finally adopted is shown in fig. 3. It consists of 



