21-1 Mr. G. F. Rodwell on the Effects 



E, and the eyes are fixed on the second hand of the watch ; at 

 a certain second the caoutchouc tubes are released at L, E is 

 simultaneously turned, and the level kept constant in D during 

 the experiment; when the half-litre flask M is full of air, the 

 caoutchouc tubes are simultaneously closed and the exact second 

 noted. By repeating- this several times, we determine the time 

 to a part of a second required by a stream of water flowing at a 

 known rate to carry down half a litre of air under the conditions 

 of the experiment. 



It is obviously necessary that the level in A should remain 

 perfectly constant during an experiment — that is to say, that the 

 influx and efflux should be equal ; for if the efflux from A be 

 greater than the influx, less than the real amount of air carried 

 down will make its appearance in M ; if, on the contrary, the 

 influx be greater than the efflux, more than the real amount of 

 air carried down will appear in M : it is also necessary to keep 

 the level in D constant, otherwise the pressure on the orifice of 

 C, and consequently on the air iri A, will var}^. It is difficult 

 to cause the descending current of water to pass exactly down 

 the axis of the tube B, and a very slight variation from that 

 position greatly diminishes the amount of air carried down ; 

 when perfectly in the axis, a peculiar roughness is added to the 

 roaring sound produced ; but in order to remove any possible 

 source of error from this cause, the amount of air carried down 

 by the stream from each delivery-tube, under certain known 

 conditions, when the stream passed absolutely down the axis of 

 B was determined, and before each set of experiments, the same 

 conditions w r ere made to obtain, and the delivery-tube adjusted 

 until the standard amount of air, previously found to be carried 

 down under those conditions, was carried down, when (the stream 

 being known to be in the exact axis of B) any other conditions, 

 as to pressure &c., not involving the removal of the delivery- 

 tube could be made to obtain. Great care was taken that the 

 whole apparatus should be perfectly vertical before each set of 

 experiments, and the whole was made as immoveable as possible. 



It may be considered unnecessary to give these details of ex- 

 periment, but I am well aware that results are often doubted 

 because their author has neglected to state the exact method by 

 which he obtained them. 



Before giving the results obtained by the apparatus described 

 above, it will be as well to consider some facts connected with a 

 water-jet. 



A jet of water moving with considerable velocity from above 

 downwards is observed to widen gradually as it gets further from 

 a point just below the orifice from which it issues; and at a cer- 

 tain distance from the orifice the velocity of the particles becomes 



