Mr. A. Bird en a Water-Barometer. 351 



tried in summer, when the external temperature is 70° or 80°, 

 the sudden cooling causes so great an evaporation from the sur- 

 face of the water, and condensation in the upper part of the baro- 

 meter, that a real rain-shower is produced, the condensed water 

 running down the glass tube in innumerable pellucid drops in 

 the most beautiful manner, thus perfectly imitating the conden- 

 sation of invisible watery vapour in the higher regions of the 

 atmosphere. When the compo tube leaves the zinc vessel, it is 

 led up perpendicular to the Lambert tap L. Above the tap L 

 the tube still rises perpendicular, when it suddenly bends down, 

 leaving the end open at M. 



I now describe the part of the barometer below the glass tube. 



The compo tube being soldered on, was carried down to the 

 cistern, not necessarily perpendicular; for instance, the tube 

 may descend at an angle of 30° or 40°, and may twist in 

 any convenient direction. The entire instrument erected by 

 me is in the house, to escape a freezing temperature. At the 

 lowest bend of the compo tube is a short upright tube, having at 

 the end a Lambert tap N, to which is soldered a male screw of a 

 fth-inch gas union-joint O, the use of which will be understood 

 further on. The compo tube now begins to ascend j and at the 

 top of the bend is another Lambert tap P. Beyond this the compo 

 tube bends down and reaches nearly to the bottom of the cistern, 

 which is a one-gallon white-glass narrow-mouth upright bottle 

 R. The bottle rests upon a stand S, which moves up and down 

 by means of a set screw T, acting through a stout shelf U U ; 

 and the bottle is kept steady by means of the two uprights W,W, 

 upon one of which is fixed the zero-point E. 



I shall now describe the method of filling the barometer, which 

 was as follows. 



Four gallons of water were carefully distilled, and being put 

 into a perfectly clean and new tin oil-can with a narrow mouth, the 

 water was boiled for one hour over a bright fire, the object being 

 to drive out the air. While still boiling, two quarts of olive oil 

 were poured in. This slightly increased the pressure in the 

 water underneath, causing the last remains of the air to rise with 

 the steam in jets or spirts through the stratum of oil. The 

 instant ebullition was stopped, the oil closed over the boiled 

 water, and it became hermetically sealed from the atmosphere. 

 The contents of the tin can were now cooled, and the can X was 

 placed above the top of the water-barometer. A piece of -|th- 

 inch gutta-percha tube Y Y, sufficiently long to reach from the 

 can X above to below the very bottom of the barometer, was pro- 

 cured, and one end of the tube was put into the mouth of the 

 can X, the end passing through the supernatant stratum of oil 

 down to the bottom of the water underneath. At the other end 



