[ 316 ] 



XXXIX. All Automatic Mercury Vacuum-pump. Bt/ S. R. 

 MiLNEK, D.Sc. (Lond.)j Lecturer in Physics^ University 

 College, Shefield*, 



[Plate Xir.] 



THE vacuum-pump described in the following pages is 

 the outcome ot* an attempt made to modify the ordinary 

 form of Topler pump into a simple and portable automatic 

 instrument of small size. It possesses an improvement in the 

 form of pump-head, and in the method of expelling the last 

 traces of air from the pump-bulb, which I believe conduces 

 to a greater efficiency in working than is the case with the 

 ordinary form, and may therefore make a description of the 

 instrument of interest. 



The pump is of the shortened type worked by a water 

 vacuum-pump, and its general form may be seen by reference 

 to fig. 1 (PI. XIL). (The separate parts are spread out in the 

 figure for the sake of clearness — in reality the whole apparatus 

 is contained on the two sides of a board 80 cms. high by 25 wide.) 

 The reservoir containing the mercury, A, is about 30 cms. 

 only below the pump-bulb B, and an electrically worked 

 three-way tap, T, connects the air space of A alternately to 

 the atmosphere and to a water vacuum-pump. This causes 

 the mercury to rise and fall in the pump-bulb, and to expel 

 the air from it into an '^ external vacuum '*' chamber E E', 

 which has been previously evacuated by the water-pump. 



The tap T is caused to turn at the proper times by means 

 of contacts, made by the mercury of the pump, actuating 

 -either of the pairs of coils K or K' (of which only one coil 

 ^ach is shown in the figure). A current through K causes 

 the lever 1 1' attached to the plug of T to be pulled down on 

 the right-hand side to the correct position for the tap to 

 make the connexion it to r. This opens A to the atmosphere 

 throuoh a drvino-tube not shown in the fioure. During the 

 motion of the lever a small screw h attached to it, by knocking 

 down the bent lever i, breaks the contact ij in series with K, 

 while it makes a contact ij in series with the other coils K'. 

 The current through K which has produced the motion is 

 thus immediately cut off, but the lever remains in position 

 through the friction of the tap, until a current passes through 

 K'; this causes a similar action on the left-hand side and 

 a connexion uv of A to the water-pump to be made. All the 

 breakings of contact thus take place outside the pump, so 

 that no sparking ever occurs in the mercury, and a further 



* Communicated by the Author. 



