54 LAWRENCE HAKGRAVE. 



This is not thought to be an extravagant number compared with 

 the generality of expansion gears. 



The drawing shows the arrangement of parts clearly and the 

 valve has only failed on one occasion when it was choked with a 

 scrap of dirt or sand. The engine works if the valve is not held 

 by either catch, but then the piston has not the full stroke. The 

 catch that holds the valve in its lower position is too high up in 

 the drawing, the steel plate into which the valve is screwed would 

 knock against the catch instead of against the head of the guide 

 bolt that regulates the travel of the valve. It will be seen that 

 one catch is fast to the valve and the other to the valve cylinder; 

 in the model exhibited both are on the valve cylinder, this does 

 not alter the principle of the valve gear, it is an adaptation to 

 suit this particular case. That thing like a string of beads 

 in the drawing is a spring | inch diameter, made of steel music 

 wire No. 8 ; it is soldered by the middle to the valve plate. Some 

 of the failures of these vibrating engines have been due to the 

 ports being too small, this is an error that might have been avoided 

 by a little more elementary knowledge of engineering. 



The receiver ends and cylinder covers are pressed, but do not 

 think for a moment that a fly-press is one of the writer's laboratory 

 tools, neither are the dies steel nor hardened. Any scrap of brass 

 or iron is used for the dies if the size is suitable, and the disc for 

 the proposed cover is cut out with the tin shears, a grip in the 

 vice completes the operation unless a ragged edge has to be turned 

 off in the lathe. The test pressure with water gives the required 

 amount of dish to the receiver ends. The piston and junk ring 

 are made of vulcanite, and the cup leather packing does away 

 with the necessity for the cylinder being either round or parallel. 



The air-pump has proved itself a serviceable tool, it has a bent 

 lever handle and long links on the principle of the Stanhope press 

 so that the most powerful leverage acts on the cylinder full of air 

 when it is reduced to its smallest volume. The valves are made 

 very small and great pains were taken to reduce the clearances to 

 a minimum. The ram is 1 1 inches diameter, and the stroke about 

 4^- inches. The suction valve is in the bottom of the ram. through 

 which the air is drawn. The delivery valve seat is ^ inch diameter 

 and there is no water jacket to the cylinder. The receiver in 

 Plates I. and II., is charged in six minutes to 230 pounds, the 

 pump is then warm. When the pump leather is in good condition 

 a pressure of 400 pounds can be shown on the gauge. Should 

 more high pressure air-compressing gear be required, a water pump 

 will be used to force water into a large receiver to compress the air 

 therein, thus avoiding the loss of power by the expansion of the 

 air left in the cylinder and passages between the valves at the 

 end of the stroke. It may be pointed out that even the air-pump 



