ON NA VAL ARCHITECTURE. 307 



enlarged to its original diameter, the water must flow through the 

 narrow portion with increased velocity that is a mere arithmetical 

 consequence. Prima facie, it seems certain that the water must be 

 very much squeezed in going through the narrow part, and must there 

 tend to push that part of the pipe forwards ; but on the contrary, the 

 moment you look into it, you will see that instead of the water being 

 more pressed at this point than anywhere else, is a great deal less 

 pressed, there is much less pressure there than anywhere else, and if 

 you have a series of vertical pipes connected with the interior of this 

 pipe and the water is allowed to rise in them until it stands at a height 

 representing the pressure, you will see that at the -part before the 

 contraction has been created, the water will rise its natural level, but 

 in the contracted part it will have been greatly lowered, and indeed 

 if the contraction is sufficient there may be produced a partial 

 vacuum. There is a little instrument which Lord Rayleigh showed me 

 the other day, in which by connecting a pipe of that sort with a' vessel of 

 mercury, and blowing through the pipe with the mouth, you are enabled 

 to raise eight, nine, or ten inches of mercury by the partial vacuum 

 that forms itself in the contracted portion. The rationale is clear. The 

 particles of water at the commencement of the contraction are going 

 at a certain speed, and in the narrow part the same particles are 

 inevitably going with a greater speed. How can particles have possibly 

 acquired increased speed in that direction except by the circumstance 

 that the pressure in front of them is less than the pressure behind them? 

 The very fact of this acceleration is a proof, in terms, that there is 

 less pressure in front than behind, and for that reason this contraction 

 is a region of small pressure, and the original pipe a region of com- 

 paratively large pressure. Vice versa, it you take a pipe and put an 

 enlargement into it, we find the opposite thing happens. There, the 

 particles which were going with a certain speed at the commencement, 

 when they arrive at the enlarged portion are going with a proportionately 

 reduced speed, and that can only have happened by there being more 

 pressure on each particle in front than behind it. Each particle as it 

 enters this region is being pushed backwards. As the pipe gradually 

 resumes its natural diameter after passing either a contraction or an 

 enlargement, the original pressure is also gradually resumed, and 

 regarded as endways forces, the alternations of pressure exactly balance 



