318 



In such states of the rails a gradient as shght as 1 in 

 160 or 150 will sometimes reduce the speed of an ordinary 

 passenger train as low as 10 miles per hour. 



In estimating the various sources of loss of power on the 

 Atmospheric system, it is ascertained both by calculation and 

 experiment, that there is practically no material loss in the 

 production of the prior exhaustion ; that the loss of power 

 by friction of the stationary engine itself will amount to about 

 six per cent., as ascertained with respect to large stationary 

 engines of the nearest similar construction. 



The loss of power by friction of the air pump I have not 

 precise means of ascertaining ; it may amount to about four 

 per cent., but most probably less, certainly not more. The 

 friction of the engine and air pump at Dalkey appear, as 

 taken by the indicator, to be about ten per cent. 



There is, as shown by the experiments at Dalkey, con- 

 siderable loss in opening the air pump valves, which, being 

 large, shut with great violence, and are consequently fixed on 

 the valve seats so as to require considerable power to open 

 them, amounting almost to ten per cent. ; but this amount 

 does not arise from the absolute power required to open the 

 valves, but because the valves being opened by the pressure 

 of the air entering into or expelled from the pump, according 

 to a well known hydrostatic law, require as much more force 

 to open them by the pressure from the piston than by direct 

 action as the area of the pump piston is larger than the area 

 of the valves. At Croydon, for the sake of avoiding the 

 concussion of the outlet valves, small valves are used of 

 about three inches in diameter, and thirty in number at each 

 end of the pump. I do not know how far this arrangement 

 is preferable to that at Dalkey, but it appears to me that the 

 valves at Croydon require very considerable power to open 

 them, and that the air is much condensed before it makes its 

 escape from the pump. 



I 



