the Mer curia I Air-pump. 135 



matter if the capacity of the bell-jar to be pumped out is at all 

 considerable in comparison with that of the exhausting-vessel B 

 of the mercurial pump. 



The ideaof the arrangement that has just beendescribedoccurred 

 to me last autumn at the Meeting of Men of Science [the annual 

 Naturforscherversammlung] at Giessen, where I had an opportu- 

 nity ot' being present at some of the experiments made by Geiss- 

 ler of Bonn with his pump, but witnessed also how easily the 

 instrument can be destroyed by unskilled hands. 



On my return to Berlin I determined to have my idea carried 

 out, at least in the way of an experiment. I had already en- 

 trusted it to a mechanician to execute, when, to my not altogether 

 agreeable surprise, I saw, from the September Number of tile 

 Philosophical Magazine, which had in the meantime arrived, that 

 the Rev. T. R. Robinson, the same to whom we are indebted for 

 a valuable investigation concerning the lines of the spectrum, had 

 already had a mercurial pump constructed upon similar principles. 



Only one half of his plan, however, is coincident with mine. 

 Like me, he uses a common air-pump to remove the mercury 

 from the exhausting-vessel ; but in order to till it he requires, 

 besides, a separate apparatus, consisting of an upright cylinder 

 of cast iron, 13 inches high and 3*2 inches wide, whence the 

 quicksilver is driven by means of a wooden plunger into the ex- 

 hausting-vessel. 



In consequence of this and other parts of the arrangement, 

 the instrument, which holds only ten pounds of mercury, becomes 

 so complicated and clumsy in form, that it is difficult to believe 

 it will meet with a very favourable reception' at the hands of phy- 

 sicists. Consequently I did not consider that I need hesitate to 

 proceed with the practical execution and publication of my idea. 



I will in conclusion also remark that, when all that is wanted 

 is to exhaust simple tubes with platinum or aluminium wires 

 melted into them, a regular mercurial pump is not at all needful, 

 even in order to attain as high oV even a higher degree of ex- 

 haustion than can be reached by means of it. 



For this purpose nothing more is needed than a small tubu- 

 lated gas-generating bottle about half filled with quicksilver, and 

 the neck and tubulus ot which are closed by good corks. Through 

 the cork of the tubulus is inserted an iron tube which can be 

 closed by a stopcock, and ends externally in a nozzle ; and through 

 the cork in the neck of the bottle is inserted a glass tube about 

 8 or 10 inches long, which must be melted on to about the 



