40 SECTION PHYSICS. 



these tubes are attached to a simple U tube, mounted with some little 

 care, such as you see here, by which the pressure is measured. As 

 the exhaustion of the chimney operates on each lime equally, the ether 

 in the U tube will not be moved by it, but as one is drawn by virtue of 

 the slight vacuum experienced, and the other pushed by virtue of the 

 slight pressure, they both act in the same direction, and so the ether 

 used in this U tube is moved, one limb being depressed and the other 

 elevated. Affixed to the flexible tubes is a small switch or button, by 

 reversing which, the limb which was first depressed is elevated, and 

 vice versa. In using the instrument the operation will be to place the 

 scales level with the surface of the ether in the first instance, then to 

 reverse the button, and then move the scale to the new position and 

 read again, and by deducting the one from the other, you get a reading 

 which is double the amount of the elevation. Thus if the difference of level 

 is Toth of an inch you get iVhSj an d this double reading of course halves 

 the error. The question then was to find out how this was to be con- 

 nected with the speed, and for that purpose I made a mathematical 

 calculation which gave this formula, V= N /pX28.55. I made a number 

 of experiments with currents of air of known x'elocity for the purpose 

 of checking this result. There was some difficulty to get currents of air 

 of known velocity, but I accomplished it by getting a flue of 100 

 feet in length, connected with a chimney letting off a small flash of gun- 

 powder at one end, and by means of a pane of glass, by which I could 

 see through the flue, I counted the number of seconds which the 

 smoke took in coming from the starting point to the place where I 

 stood at 100 feet distance. Estimating the exact speed in that way, 

 and testing it with the instrument, I got the constant 28.50., being 

 only 5 from the figure given by calculation. I then calculated a table 

 accordingly. I also made a table of corrections for temperatures. 

 The first table gives the speed, supposing the air to be at 60, but as 

 the air in these flues is often as high as 500, or 1000, that also 

 required a careful calculation, which I have given. I take this 

 opportunity of shewing this instrument, thinking it may be useful 

 to those who have to measure gases where ordinary Anemometers of 

 the mechanical kind cannot be introduced. In this there is no friction 

 whatever, as there are no moving parts. The column of liquid in the 

 Anemometer attains its position, and is there kept by the constant 



