134 Prof. Poggendorff on a new form of 



It requres, as is easily seen, no great expenditure of strength, 

 and is moreover not attended with any danger; besides which, 

 the apparatus is placed in a strong wooden vessel, large enough 

 to hold the whole quantity of quicksilver in the scarcely con- 

 ceivable case of an accident. 



Every repetition of the operation is begun by separating the 

 bottle A from the pump, and, by carefully turning the cock/, 

 allowing the quicksilver to mount slowly up into the vessel B. 

 In consequence of the usually small capacity of the pieces to be 

 exhausted, this vessel fills again almost completely of itself, so 

 that, after opening the cock g, only a small quantity of air has to 

 be removed from below the upper nozzle. 



It is obvious that this apparatus, when carefully constructed, is 

 capable of performing the same services as Geissler's pump, and 

 that in fact it differs from this only in the way in which the 

 emptying of the mercury is effected. 



I will here mention only one of the preliminary experiments 

 which I have made with it, — an experiment which is calculated 

 to refute the common opinion that an absolute vacuum is to be 

 attained by means of the mercurial pump. This consisted in 

 trying whether the vessel B, after being emptied of quicksilver, 

 would allow a current of electricity to pass. This I found to be 

 really the case. If the iron caps of the apparatus are connected 

 with the poles of an induction-coil, an iron wire being put 

 through the cock /into the quicksilver in the bottle A, the welU 

 known luminous phenomena are obtained most beautifully de- 

 veloped. 



As already remarked, the apparatus that has been described is 

 specially applicable for exhausting small vessels which can be 

 directly connected with the lateral opening /. It can, how- 

 ever, be employed also for pumping out larger vessels, such as 

 bell-jars which require to be placed upon a plate. 



This is effected by putting the side opening /, by means of a 

 flexible metallic tube, into connexion with the air-pump through 

 the cock which, in Pistor's air-pumps, leads to the Hawksbee's 

 barometer-gauge, and is provided for this purpose with a second 

 perforation, similar to that in the main-cock h. 



By first of all placing the cock h so as to establish a commu- 

 nication from the pump-barrels to the bell-jar standing upon the 

 plate, this can be exhausted as far as the pump allows ; and by 

 then turning the main-cock through a right angle, and connecting 

 it by the caoutchouc tube with the mercurial apparatus, the ex- 

 haustion can be continued by means of the operations previously 

 described, so as to produce a more perfect vacuum. 



This, however, of course requires that all the joints should be 

 perfectly air-tight ; and even in that case it becomes a tedious 



