308 FLOW THBOUGH TUBES OF UNEQUAL BORE. 



have to be increased. Now, in this case, the velocity o1 

 the liquid remains the same when it reaches the wider 

 part, as a consequence of the inertia of the liquid 

 particles, the jet therefore maintains its previous 

 dimensions in width, and does not fill the wider portior 

 of the tube. 



The case becomes somewhat different, when the flo^ 

 takes place in air. Left to itself, the efflux of the liquic 

 takes place in precisely the same way as shown in fig. 20$ 

 4- and .5, for the efflux in vacuo. But if the orifice c 

 be now closed for a short while, and then opened again 

 a jet will issue, having the full dimensions in width o: 

 the orifice, fig. 202 C. It is the pressure of the externa 

 air which now prevents the partial emptying of th< 

 wider portion ; for if the issuing jet is to have the 

 greater width of the full orifice and the water is to flov 

 out with the same velocity which the particles possess 

 the liquid would be expected to break up into detachec 

 portions, between b and c, fig. 202 D, but the vacuoui 

 spaces which would thus arise, are rendered impossibL 

 by the pressure of the external air at a. The wate: 

 issuing now in a thicker jet, must have a diminishec 

 velocity; while on the other hand the liquid particle, 

 tend to maintain their previous velocity, and to separat< 

 from one another between b and c. The consequent 

 of this is that the pressure between b and c is consider, 

 ably diminished, and this decrease of pressure may b< 

 so great, with a suitable form of the tube, and a sufficien 

 velocity of efflux, that the pressure within the tube i 

 in that portion actually less than that at a where th< 

 liquid is discharged. The pressure at a must of coursi 

 be greater than the atmospheric pressure, or no liquk 



