VOL. XVI.] PHILOSOPHICAL TRANSACTIONS. 335 



of bodies are to each other, as the square roots of the heights to which they 

 may ascend : and so in this case they are also reciprocally as the roots of the 

 specific gravities. 



If therefote we would know what is the velocity of air when driven by any 

 degree of pressure whatever, we need only find what would be the velocity of 

 water under the same pressure, and then take the square roots of the specific 

 gravities of these two fluids, because, as much as the square root of the specific 

 gravity of water exceeds the square root of the specific gravity of air, so much 

 in proportion will the velocity of air exceed the velocity of water. For example, 

 when I would compute what would be the velocity of a bullet shot by the pneu- 

 matic engine, described in Philosophical Transactions, N° 179) I should first 

 compute what was the velocity of the air itself that drove the bullet : I there- 

 fore observe that on this occasion the air sustains a pressure much about the 

 same as that of water when its head is 32 * feet high ; now such water would 

 spout out with a sufficient velocity to ascend 32 feet perpendicular, and there- 

 fore it has the velocity of 45 feet in a second. It remains therefore only to 

 know the proportion of the gravity of air to that of water: this we have found 

 not to be always the same ; because the height, the heat, and the moisture of 

 the atmosphere are variable ; yet we may say in general, that the ratio of the 

 specific gravities of water and air is about 840 to 1 . Taking then their square 

 roots, which are 2Q and 1, we may conclude that the velocity of air must exceed 

 that of water by 29 times ; and so multiplying 45, the velocity of water, by 29, 

 we shall find that the velocity of the air, driven by the whole pressure of the 

 atmosphere, is about 1305 feet in a second. 



Extract of a Letter from Mr. J. Flamsteed, Astr. Reg. and Reg. Soc. S. giving 

 his Calculation oj" the Eclipses of Jupiter s Satellites Jbr the Year 1687. ff^th 

 a Table of the Parallaxes of the Orbit, and an Ephemeris of Jupiter s Geo- 

 centric Place for the same Year. To which is added an Observation of the 

 Eclipse of the Moon, Nov. 30, l685, made at Lisbon, and Mr. Flamsteed's 

 own Observation of the Eclipse of Jupiter by the Moon, on March 31, iQsO. 

 N° 184, p. 196. 

 The calculated times of the eclipses and of the planet's places cannot be of 



any use now. But the observations of the lunar eclipse and of Jupiter's occul- 



tation by the moon are as follow : — 



* This number is too small, as the mean pressure of the atmosphere is now known to be about that 

 of a column of 33| feet of water, and this will give for its velocity 46"~? feet, instead of 45 ; which 

 multiplied by the 29 gives near 1348 feet, for the more correct velocity with which air will rush into 

 a vacuum. 



