produced by a Fluid in Motion. 



219 



variation in the quantity of air carried down being observable in 

 a number of consecutive experiments, In experiments 21, 22, 

 and 23 the resistance was insufficient to keep a column of water 

 suspended in B, and the disk action returned, the air escaping 

 from C very irregularly. In experiment 24 there was persistent 

 collision at about 11 inches from the orifice of H, and we get 

 nearly the same result as in experiment 20. In experiments 

 25, 26, 27, and 28 the collision was persistent at 10 inches from 

 the orifice of H, in experiment 29 at 9 inches, and in experi- 

 ment 30 at 8 inches. We see, therefore, that in the above 30 

 experiments, air was carried down by disk action in 22 (viz. expe- 

 riments 1 to 19, and 21 to 23) ; and by direct collision, accord- 

 ing to the theory of Magnus, in 8 (viz. experiments 20 and 24 

 to 30). 



The nearer the collision approached the orifice from which the 

 jet issued, the less was the quantity of air carried down j and 

 when the water in B touched the orifice of H, air only entered 

 at intervals, and was seen to force its way through the water 

 into the line of the descending jet. This latter effect is due to the 

 diminution of pressure in the line of a moving fluid, — an effect 

 which may be shown by introducing a small mercury gauge into 

 the line of a current of liquid flowing in any direction ; or in a 

 more striking manner if we allow a liquid to flow from a delivery- 

 tube A (fig. 3), dipping beneath the surface of water in a vessel B, 

 and place in the line of the current the lower orifice of a tube, C, 

 open at both ends. Air will now be found to force its way into 

 the current against the pressure of the column of water D E. 



Experiments with delivery-tubes of various shapes, a constant 

 depth of 1 inch of water being maintained in the vessel D. 

 Orifice of the tube delivering the air (C, fig. 1) = -^ths of an 

 inch diameter. The comparative dimensions of the tube B 

 (fig. 1) and the delivery-tubes employed are shown in fig. 2. 



I. Circular delivery-tube T y^tha of an inch in diameter (B, fig. 2). 

 Flow = half a litre of water in 39 seconds. 



Level of water maintained constant 

 in A during the experiment. 



Quantity of air carried down by 

 half a litre of w r ater. 



At 1 inch below the orifice of B 



(Mg. 1). 



At the orifice of B (c, fig. 1). 

 At 1 inch above the orifice of B 

 (d, fig. 1). 



No air was carried down. 



211 cubic centims. 

 147 „ 



No air was carried down in the first experiment, because there 

 was not sufficient water in the stream to allow of the formation 



Q2 



