ON HYDRAULIC PRESSURE. 233 



Garnerin's parachute,* with its whole load, was about a quarter of a 

 pound for each square foot, the square of its greatest velocity must, there- 

 fore, have been about 125, and the velocity 11 feet in a second, which is no 

 greater than that with which a person would ascend, in leaping from a 

 height of two feet, without stooping. Mr. Garnerin found the velocity 

 even less than this, and it is not improbable that the concave form of the 

 parachute might considerably increase the resistance. Thus, Mr. Edge- 

 worth found that a plate 9 inches long, when bent into an arc of which 

 the chord was 7, had the resistance increased more than one seventh, t 

 The diminution of the resistance of the air by the obliquity of the surface is 

 still less than that of the resistance of water : thus, the resistance on the 

 oblique surfaces of a wedge is not quite so much less than the resistance 

 on its base, as its breadth is less than the length of those surfaces. 



When the velocity of a body moving through an elastic fluid is very 

 great, the resistance is increased in a much greater proportion than the 

 square of the velocity : thus the retardation of a cannon ball moving with 

 a velocity of 1000 feet in a second, or a little more, becomes suddenly 

 much greater than the calculation indicates. The reason of this change 

 appears to be, that the condensation of the air before the ball is necessarily 

 confined to a smaller portion which is very intensely compressed, because 

 the effect of the impulse can only spread through the air with a certain 

 velocity which is not much greater than that of the ball ; and this smaller 

 portion of air must necessarily be much more condensed than a larger 

 portion would have been. Thus, when a cannon ball moves slowly, its 

 effect at any instant is in some degree divided throughout all that part of 

 the atmosphere which the sound of the report has reached ; and if the ball 

 follows the sound very speedily, it is obvious that the portion of the air 

 before the ball which partakes of the effect, must be very small. The 

 sound is observed to be propagated with a velocity of about 1130 feet in a 

 second, and a cannon ball may be discharged with a velocity of 2000 ; but 

 one half of this is very speedily lost, so as to be wholly useless with regard 

 to the effect of the ball. If, therefore, we wish to increase the range of a 

 cannon ball, we must increase its weight ; for the resistance increases only 

 in proportion to the surface of the ball, while the weight is determined by 

 its solid content. 



It is not easy to explain, in a manner perfectly satisfactory, the reflection 

 of a cannon ball, or of a stone, which strikes the surface of the sea, or of 

 a piece of water, in an oblique direction. We may, however, assign some 

 causes which appear to be materially concerned in this effect. In the first 

 place the surface of the water, acting at first for some time on the lower 

 part of the ball, produces, by its friction, a degree of rotatory motion, by 

 means of which the ball, as it proceeds, acts upon the mass of water which 

 is heaped up before it, and is obliged by a similar friction to roll upwards, 

 so that it mounts again to a much greater height than it could possibly 



.* Nich. Jour. i. 523, 8vo. iii. 57. Gilbert's Jour. xvi. 156, 164, 257. See the 

 article Aeronautics, Supp. to Encyc. Brit, 

 f Ph. Tr. 1783, kxiii. 136. 



