568 GAS, COAL 



F, o, H, are three glass tubes, which aro firmly cemented into the caps e, e, e. F and 

 H, are each from 15 to 20 millimeters internal diameter, and are selected of as nearly 

 the same bore as possible, to avoid a difference of capillary action. The tube o is 

 somewhat wider, and may be continued to any convenient height above the cylinder. 

 H is accurately graduated with a millimeter scale, and is furnished at top with a 

 small funnel i, into the neck of which a glass stopper, about 2 millimeters in diameter, 

 is carefully ground. The tube F terminates at its upper extremity in the capillary 

 tube Jc t which is carefully cemented into the small stoel stopcock /. F has also fused 

 into it at m, two platinum wires, for the passage of the electric spark. After this tube 

 has been firmly cemented into tho cap e, its internal volume is accurately divided 

 into 10 perfectly equal parts, which is effected without difficulty by first filling it with 

 mercury from tho supply-tube a, up to its junction with the capillary attachment, and 

 then allowing tho mercury to run off through the nozzle h until tho highest point of 

 its convex surface stands at tho division 10, previously made so as exactly to coincide 

 with tho zero of the millimeter scale on H ; the weight of the mercury thus run off is 

 carefully determined, and the tube is again filled as before, and divided into 10 equal 

 parts, by allowing the mercury to run off in successive tenths of the entire weight, and 

 marking tho height of tho convexity after each abstraction of metal. By using the 

 proper precautions with regard to temperature, &c., an exceedingly accurate calibration 

 can, in this way, be accomplished. 



Tho absorption tube i is supported by the clamp n, and connected with the 

 capillary tube k t by tho stopcock and junction-piece II' p, as shown in tho figure. 

 When tho instrument is thus far complete, it is requisite to ascertain the height of 

 each of the nine upper divisions on tho tube, above tho lowest or tenth division. 

 This is very accurately effected in a few minutes by carefully levelling tho instrument, 

 filling the tube o with mercury, opening the cock /, and the stoppered funnel ', and 

 placing the cock / in such a position as to cause the tubes F H to communicate with 

 the supply-tube G. On now slightly turning the cock g, the mercury will slowly rise 

 in each of the tubes F and H ; when its convex surface exactly coincides with the 

 ninth division on F, the influx of metal is stopped, and its height in H accurately 

 observed ; as tho tenth division on F corresponds with the zero of tho scale upon H, it 

 is obvious that the number thus read off is the height of tho ninth division above 

 the zero point. A similar observation for each of the other divisions upon F com- 

 pletes the instrument. 1 



Before using the apparatus, the large cylinder D D is filled with water, and tho 

 internal walls of the tubes F and H are, once for all, moistened with distilled water, 

 by the introduction of a few drops into each, through the stopcock /, and tho 

 stoppered funnel i. The three tubes being then placed in comnrnnication with each 

 other, mercury is poured into G until it rises into the cup i, the stopper of which is 

 then firmly closed. When the mercury begins to flow from I, that cock is also 

 closed. The tubes F and H are now apparently filled with mercury, but a minute 

 and imperceptible film of air still exists between tho metal and glass ; this is effec- 

 tually got rid of by connecting F and H with the exit tube h, and allowing the 

 mercury to flow out until a vacuum of several inches in length has been produced in 

 both tubes ; on allowing tho instrument to remain thus for an hour, the whole of tho 

 film of air above mentioned will diffuse itself into tho vacuum, to bo filled up from tho 

 supply-tube G. These bubbles aro of course easily expelled on momentarily opening 

 the cock I and the stopper i, whilst G is full of mercury. The absorption-tube i being 

 then filled with quicksilver, and attached to I by the screw-clamp, tho instrument is 

 ready for use. 



In illustration of tho manner of using tho apparatus, a complete description of an 

 analysis of coal-gas by this instrument will be given below. 



For the analysis of purified coal-gas by means of tho mercury trough and eudio- 

 meter, the following operations aro necessary : 



I. Estimation of Carbonic Acid. 



A few cubic inches of the gas aro introduced into a short eudiometer, moistened as 

 above described ; tho volume is accurately noted, with tho proper corrections, and a 

 bullet of caustic potash is then passed up through tho mercury into tho gas ; it is 

 allowed to remain for at least one hour ; tho volume of tho gas being again ascer- 

 tained and subtracted from tho first volume gives the amount of carbonic acid which 

 has boon absorbed by the potash. 



II. Estimation of Oxygen. 



This gas can bo very accurately estimated by Liobig's method, which depends upon 

 tho rapid absorption of oxygen by an alkaline solution of pyrogallic acid. To apply 



1 This instrument may be obtained from Mr, Oertlicg, philosophical instrument maker. 



