

MAGDEBURG HEMISPHERES 279 



jove the mercury in the tube could be rendered a 

 erfect vacuum. 



Tubes and vessels with narrow necks may be exhausted by con- 

 acting them with the tube a and using the tube which passes 

 [rough the plate for communicating with the external air. In 

 at case the stop-cock will have to be placed in position III. before 

 e piston is drawn back, and in position II. before it is pushed in ; 

 :;at is, the order of turning the stop-cock will be simply the 

 Averse of what it is when a receiver is exhausted. For the pre- 

 nt experiment a glass tube, 90 cm long and from 2 to 5 mm wide, 

 bent at right angles 5 cm from one end, and is then connected by 

 short piece of india-rubber tubing with the tube a, the long 

 iwtion of the tube being allowed to hang straight down by the 

 de of the table. The ends of the tube a and of the glass tube 

 lust be in close contact within the india-rubber connection, or 

 e portion of the soft india-rubber tube between both ends would 

 ; compressed by the external air when the exhaustion inside the 

 be proceeds, and the proper connection of the parts would be 

 ifcerrupted. The india-rubber tube should be tied by thread 

 und both tubes, if it does not fit perfectly tight. The other 

 i.d of the tube is dipped into the vessel represented in fig. 162, 

 iled with mercury, and it should be allowed to rest upon the 

 httom of the vessel, for when partially filled with mercury the 

 ibe becomes heavy and exerts too strong a pull at the india- 

 ibber, unless it is supported below. The space to be exhausted is 

 I this experiment very small, and at the first stroke of the piston 

 te rarefaction is already so great as to raise the mercury in the 

 the to a height of about 70 cm . The quantity of mercury contained 

 i the vessel should therefore be sufficient to cover the lower 

 < erture of the immersed tube to the end of the experiment, as 

 <|herwise, if it be exposed to the air, the atmospheric pressure 

 trees the column in the tube upwards and throws it violently 

 iito the pump. When the experiment is finished, the stop-cock 

 :|ould be very slowly turned into . position I., so that the mercury 

 :| the tube may sink gradually, and not be thrown beyond the rim 

 *: the vessel. 



The great force of the atmospheric pressure is shown 



a striking manner by the apparatus called the 



fagdeburg hemispheres. It consists of two hollow 



3mispheres of metal, with evenly ground edges which 



