94 CHEMICAL DISCOVERY AND INVENTION 



in Sir William Crookes's laboratory. It will be seen that there 

 are four fall tubes, down which the mercury drops as it is 

 supplied from the reservoir at the side when the latter is lifted 

 to the proper height. There is a vertical tube placed alongside 

 the fall tubes, and this standing in a vessel of mercury acts as a 

 barometer, and so indicates by the height of the mercury the 

 first stages of the exhaustion. The apparatus shown includes a 

 Pliicker tube in which gas under examination can be illuminated 

 by the electric discharge, and the light viewed through a spectro- 

 scope, of which the end bearing the slit appears on the left. 



Another form of mercury pump was devised by Topler, and 

 is frequently used for laboratory purposes. A cylindrical 

 vertical vessel is connected at its upper end with an erect straight 

 glass tube, longer than a barometer tube, and dipping into a 

 basin of mercury. The lower end of the cylinder has a branch 

 which is connected with the vessel to be exhausted, and also, by 

 means of a flexible rubber tube with a reservoir of mercury. 

 When the reservoir is raised the mercury rises into the cylinder 

 and cuts off connection with the vessel from which the air is to 

 be withdrawn. On continuing to raise the reservoir the mercury 

 rises and drives all the air out of the cylinder into the barometer 

 tube, expelling the excess of it from the bottom of that tube. 

 On lowering the reservoir again the mercury retreats, but the 

 only air which can take its place in the cylinder is that which is 

 drawn from the vessel to be exhausted. The mercury thus 

 plays the part of a piston which moves up and down in the 

 cylinder, and by repeating the operation a sufficient number of 

 times the air within can be expanded indefinitely, and ultimately 

 a very good vacuum can be obtained. 



The degree of exhaustion obtainable by any of the pumps 

 mentioned is dependent not only on the efficiency oi construction, 

 but on the vapour pressure of the liquid used at the temperature 

 of the air. Using water there is not only the vapour pressure of 

 water, amounting to between 12 and 15 millimetres of mercury 

 at room temperatures, but the atmospheric gases held in solution 

 by all ordinary water supply. Hence the exhaustion obtainable 

 by a water pump is always far from complete, although it may 

 be amply sufficient for the majority of laboratory operations, 

 such as distillation under diminished pressure. 



Even mercury gives an appreciable vapour pressure at room 

 temperatures (at 20 C. it is -001 mm.), and hence the efficiency 



