Adeney and Becker — Solution of Nitrogen and Oxygen. 393 



replaced, and the tap closed, thus leaving the tube completely filled with 

 water. 



The tube was then clamped vertically, and the lower stopcock connected 

 with a standard burette by means of pressure tubing full of water. The level 

 of the burette having been read, both stopcocks of the tube were opened, and 

 the water was allowed to flow from the tube into the burette until the 

 correct volume of water (15 cc. in most cases) had flowed out. All the 

 taps were then closed, and the pressure in the bubble adjusted to that of 

 atmospheric by repeatedly turning the two-way tap so as to connect the 

 bubble alternately with the atmosphere and the manometer. 



The tube was then swung steadily round a central axis, so as to invert it, 

 and allow the bubble to form without splashing ; when the bubble reached 

 the inverted end of the tube, the latter was returned to its first position, and 

 on the bubble reaching the top, the pressure in it was measured by opening 

 connexion with the manometer. The time taken by the bubble to traverse 

 the tube was measured in every case, and found to be practically constant, 

 at eighteen seconds for a double journey. 



The air in the bubble was renewed after each observation by taking out 

 the stopper and inserting a tube connected with a filter-pump, which drew a 

 current of air through the space. The stopper was marked, so that it could 

 be always inserted in the same position. 



As the water approached saturation and the readings of the manometer 

 got smaller, the number of inversions between each reading was increased 

 gradually from two up to ten. The readings were continued until after ten 

 inversions the manometer showed no distinct movement ; and the water in 

 the tube was then drawn into the pump and analyzed for dissolved gases. 



Considerable difficulty was at first experienced in getting an experiment 

 done at an absolutely constant temperature, owing to the fact that the tempe- 

 rature of the laboratory varied considerably, and, when the tap-water was 

 tested, it was also found to be subject to rapid and relatively large variations 

 in temperature. 



The difficulty was overcome by making the manometric observations in a 

 room the temperature of which was not subject to serious fluctuations. A 

 supply of water at steady temperature was secured by setting up in this room 

 a large thick-walled wooden vat capable of holding about sixty gallons, and 

 filling this with tap- water, the day before it was required for use, thus allow- 

 ing it to attain a steady temperature overnight. During the course of an 

 experiment, the water was allowed to syphon over through the water-jacket 

 of the apparatus continually. In this way it was found possible to keep the 

 temperature of the tube constant within 01° C. during an experiment. 



SOIENT. PKOC. R.D.S., VOL. XV., NO. XXXI. 3 R 



