430 Prof. R. Bunsen's Spectral- Analytical Researches. 



of an intercalated Ley den jar, is allowed to pass horizontally 

 across the perpendicular slit ; the length of the spark between 

 two thick platinum points is from 1 to 2 ccntims. 



The carbon points, for the reception of the liquids to be tested, 

 are prepared from the ordinary, but not too porous, drawing- 

 charcoal. In order to render them conductors, a number of 

 small charcoal rods are placed in a covered porcelain crucible, 

 which is set in a larger clay crucible and completely surrounded 

 with charcoal powder ; the crucibles are then exposed to the most 

 intense white heat for a considerable time. The small rods are 

 now pointed by means of a lead-pencil-sharper ; and the little 

 cones thus formed are cut off with a fine watchmaker's saw. A 

 supply of these cones sufficient for many years may be prepared 

 without difficulty, inasmuch as one worker can easily make ready 

 five hundred in one day. The silica, magnesia, manganese, iron, 

 potassium, sodium, and lithium which may be contained in the 

 points require to be removed ; for this purpose about a thousand 

 of the points are repeatedly boiled in a platinum dish, first with 

 hydrofluoric acid, then with concentrated sulphuric acid followed 

 by concentrated nitric acid, and finally with hydrochloric acid, 

 each acid being successively removed by boiling with water. The 

 carbon points are ready for use as soon as a hole, of the same 

 size as the platinum supports, has been bored in the base of each 

 by means of a fine three-edged borer. For each experiment new 

 carbon points are placed on the platinum supports. The liquid 

 to be examined is brought on to the points by means of a small 

 capillary glass tube warmed, if necessary, by a small gas-flame. 



A carbon cone weighs about 0*015 grm., andean absorb more 

 than its own weight of liquid. The spark-spectra obtained with 

 these cones are of long duration ; so that, if the cone be thoroughly 

 impregnated, a renewal of the liquid by means of the small ca- 

 pillary tube will not be necessary for a considerable time. 



Carbon points saturated with liquids giving simple spark- 

 spectra, and also those liquids themselves, should be arranged in 

 special small glasses, so that the normal spectra may be available 

 at any time for comparison. 



The spectrum-apparatus which had served for our earlier ob- 

 servations* was used for the production of the spectra in these 

 researches. A small mica plate which could be easily removed 

 and cleaned was used as a protection to prevent particles of liquid 

 from being carried into the slit. 



[To be continued.] 



* Phil. Mag. S. 4. vol. xxii. p. 505. 



