39 
friction of the issuing stream of particles, so that the amount 
of discharge if^ only ‘62 of that required by theory ; so from 
the varied mobility of different gases there was an antece- 
dent probability that an ideal law would not prevail for the 
velocity with which air has been assumed to flow into a 
vacuum. Hence, just as the hydraulic co-efflcient *62, 
expressing the actual amount of efflux through a hole in a 
thin plate, could only be arrived at by experiment; so by 
experiment only, could the actual velocity with which the 
atmosphere rushes into a vacuum be ascertained. This 
velocity, therefore, as determined by experiment, may be 
represented by the co-efficient ‘77 for the contracted vein. 
Or y = *77 X- 1332 = 1 025 feet per second. 
From Tables I. and II. it will be seen that the corollary 
of the equality of the velocities for all pressures, when air 
flows into a vacuum, is not strictly applicable for the lower 
pressures, but is approximately true for pressures above 
1201bs. 
That air of lower density acts as a vacuum to the 
discharge into it of air of higher density, under certain 
conditions, is a truth so well established from the experi- 
ments described as to require no further proof, but, that the 
reduction of temperature at the orifice of the discharging 
vessel did not sensibly affect the velocity of the air through 
the orifice under such conditions, will be seen from an 
inspection of the tables, and more particularly of Table Y., 
where a pressure of six atmospheres acts as a vacuum to a 
pressure of 9 atmospheres. In this experiment it will also be 
seen that 21*22 cubic inches of air, of a constant density of 
9 atmospheres, (the equivalent of 51bs. of pressure) were 
discharged successively into a vacuum and into atmospheres 
of increasing densities up to 6 atmospheres, when the 
several discharges were made in equal times, viz. 7*5 seconds. 
Now, the velocity for this time, as shown in Table L, is 
1210 feet per second for the contracted vein, and as the 
