20 Prof. Magnus on the Motion of Fluids. 



itself until by increasing the distance the difference of pressure 

 vanishes. 



42. In like manner it is easy to explain why, when two plates 

 are placed facing each other under water, and only a short 

 distance apart, and when a jet issuing with some force through 

 the middle of one of them strikes against the other, they 

 move nearer together, as s1io\to by Hatchette ; and likewise why 

 a pair of similar plates in the open air approach when an air- 

 current passing with some velocity through the centre of the one 

 meets the other. 



43. The whirling motions noticed in § 30, and represented in 

 tig. 11, are a necessary consequence of the lateral motion im- 

 parted to the water by a strong jet. When the orifice is not too 

 small and is tolerably expanded vertically, with a sufficient velo- 

 city, funnel-shaped concavities are formed ; for the water within 

 these whirlpools continually streams towards their outer rim, 

 and thus the pressure at the centre is decreased. 



44. An exactly similar action takes place when the moving 

 body is gaseous. It will be sufficient to call to mind the pecu- 

 liar motion of ascending smoke. But, if I mistake not, the spiral 

 motions observed sometimes in stormy weather are caused by a 

 similar streaming of the air towards that already in motion. Even 

 that remarkable phsenomenon the waterspout is capable of easy 

 explanation on the same principle ; for if the wind whirl with 

 sufficient velocity at some height above the Avater, the air at the 

 centre of the whirl will stream towards that outside ; in this 

 way the atmospheric pressure in the interior is lessened, and the 

 water is consequently forced up, partaking as it ascends of the 

 rotary motion of the air. Many other phsenomena might be 

 cited which are capable of explanation by reference to this lateral 

 motion, but the foregoing will, it is trusted, be found sufiicient. 



Appendix. — On the Water-Bellows. 



In conclusion, I must not omit to introduce a piece of mecha- 

 nism, the principle of which is to be referred to the penetration 

 of a fluid by air, as already described in § 15 to § 22. This is the 

 so-called trompe, or water-bellows, which, according to Grignon *, 

 was discovered in Italy about the year 1640. Such an instru- 

 ment is represented in fig. 13, drawn from the description given 

 by Richard in his Etudes sur VArt d'extrah'e immediatement 

 le Fer de ses Miner ais, p. 169. 



B is a water-holder, kept always full by means of the canal Z. 



AA are two tubes or hollowed beams, about 13 feet long, one 

 of which is here shown in section. 



* Mmoires de Physique, p. 196. 



