8. R. Williams — New Type of Mercury Air- Pump. 13 



Art. II. — A New Type of Mercury Air-Pump ; by 

 S. R. Williams. 



Perhaps no single piece of apparatus is more generally used 

 in a physics laboratory than an air-pump. Demonstrations 

 involving its use are to be found in mechanics, in heat, and in 

 all the other branches of physics, especially in that compara- 

 tively new and widening field of study, the discharge of 

 electricity through gases. A still larger field of usefulness 

 presents itself in the manufacture of incandescent electric 

 lights. 



Of the many and varied types of air-pumps each has its own 

 advantages and disadvantages, but for efficiency we recognize 

 two conditions as essential. These are, first, the power to 

 exhaust rapidly and to a high degree, and, secondly, the ability 

 to exhaust against atmospheric pressure. Other desirable fea- 

 tures are, of course, compactness, simplicity, and cheapness of 

 construction. 



Among the various types of air-exhausting pumps the Geryk, 

 or pulsometer pump (Fleuss Patent), comes the nearest to 

 fulfilling the requirements asked for above. It exhausts rapidly 

 and against atmospheric pressure, but the degree of exhaustion 

 cannot go beyond the vapor pressure of the oil used, and for 

 the average pulsometer which one finds in laboratories through- 

 out the country the degree of exhaustion cannot go beyond 

 the vapor pressure of the moisture which has been drawn into 

 the pump and mixed with the oil. In other words, too much 

 care and trouble must be taken with the pulsometer to keep it 

 in good working condition. It ought never to be used in an 

 experiment to show the triple point of water. 



Among the mercurial pumps those who have used the 

 Sprengel or Toepler, or any of their numerous modifications to 

 any great extent, must have welcomed the advent of the Kauf- 

 mann, or the Gaede pump, which could be driven with a 

 motor, and at the same time worked rapidly. The Kauf- 

 mann and Gaede pumps, however, will not exhaust against 

 atmospheric pressure. This means hitching another pump in 

 tandem with them, which is inconvenient. Frequently this 

 preliminary pumping is clone by means of a filter pump, but 

 in many laboratories the water pressure is not sufficient to do 

 this. 



It is interesting to note that in the best types of air-pumps 

 mentioned we have their prototypes in devices for lifting 

 water. The piston pump corresponds to the pulsometer, and 

 the old spiral of Archimedes was the forerunner of the Kauf- 



