G A 2

G A 2

( 532 )

increaiing the weight of the counterpoife inproportion as the cylinder rifes from the water.

Fig. 63 is a gazometer on the fame principleas Fig. 61 & 2, and differing from it only inthis Angle circumftance, viz. that ihe outer pailhas the greater part of its capacity filled by ahollow cylinder h. h. which is clofed on allfides except a perforation through its center toadmit the upright air-tube g, and is firmly fol-dered to the bottom of the pail. The ufe of thisfixed cylinder h is to leflen the bulk of waternecefi'ary to confine the gas in the moveablecylinder b, and which in the firft defcribedgazometer is often inconvenient from its weight,as it fills the entire capacity of the pail exceptthe fmall fpace occupied by the mere fubftanceof that part of the cylinder b which is immerfed.Whereas in Fig. 63 the quantity of water re-quired is only that which will fill the fpacea a between the pail and the fixed cylinder h. h.in which fpace the gas-cylinder b moves up anddown freely. The vaulted form of the upperpart of this latter cylinder is alfo an improve-ment, as it admits of the air remaining in itbeing more completely emptied by the'pipe gbefore a frefh diftillation of gas is begun.

Plate. VI. Fig. 64 is a very ingenious andufeful gas-holder, with a ciftqm at the upperpart, which on the whole combines more con-veniences in the management of gaffes than anyother. It confifts of a tinned iron hollow cy-linder a highly japanned, on the top of whichis fupported by three folid legs h. h. the pan orciftern b. (the glafs jar c is not a part of thisapparatus, but is only ihewn in this place asconvenient for many experiments). The hol-low cylinder a is plain in the infide, except afmall projection at the bottom of d. It has thefollowing fix openings into it. From the topproceed the pipes f & g, each of which alfocommunicates leparately with the pan b. (thetube g is often made parallel with f, whichanfwers as well.) The fhort tube e alfo opensdirectly into the upper part of the cylinder, dis a pretty large opening formed by a fhort pipe- which enters the cylinder at an angle of about4.5°. and paffes down for an inch or two in thefame direction. At this angle the water will notrun out through it from the cylinder, providedall the tubes that would admit the preffure ofthe external air are clofed. This opening d,has a {topper which may be ferewed on whenwanted. Befides thefe four openings, whichare effential to the apparatus, there are twoothe.rs, k. k. one at top and the other at bottomthat communicate folely with tire graduated

glafs tube z, which is ftrongly foldered into i. I*which are of brafs. The ufe of this is to fhew-by infpefition of the glafs tube the level of thewater within the cylinder, and-confequently thequantity of gas which it contains. The uiualcapacity of the cylinder for experimental pur-pofes is about 12 quarts, which are marked on apaper fcale cemented to one fideof the glais tube.

To fill this apparatus, firft hop the openingsd and e, and openy and g, then pour water intothe pan b, which will fall down into the cylin-der by one of the open tubes, whilft the airwhich it contained will efcape by the other.When the cylinder a is quite full of water,which is known by infpecling the glafs tube,flop the cocks at f and g, and open d , and’the water (as already mentioned) will not Sow-out. Then introduce the beak of the retortor other conducing tube, fo deep into d, thatit projefts beyond the inner edge, and beginthe production of the gas. As this rifes intoa, it difplaces an equal quantity of water, whichnow efcapes through d, by the fide of the tubethat brings the gas. When the gazometer is fofull that the level of the water falls to the bot-tom of the glafs fcale, fhut the tube d, and themachine may then be conveyed any where with-out danger of lofing any of the gas.

This apparatus will anfwer a variety of pur-pofes. If, for example, a jar or any otherveffel is to be filled with the gas, firft fillthe jar with water and invert it in the pan b,which is alfo to be filled. Then open the cocksat f and g , and the gas will rife into the jar byone tube and the water fink into the cylinderby another. An empty bladder may be filledwith the gas in the fame manner by ferew-ing or otherwife fattening the neck upon oneopening and letting water fall down the other jor it may be more conveniently fattened to thefide opening e , and then only one ofi.the uppercocks mull be opened to let water down. Theopening c, is alfo very convenient for fupplyinga blow-pipe, and for this purpofe the flexibletube Fig. 68 may be faftened upon e. The glafsveffel c furnifhes a convenient method of fhew-the deflagration of iron, phofphorus, &c. inoxygen gas without ritk to the veffel. It isquite open at the bottom, and has an openingof moderate fize at the top, to which is fitted acork furnilhed with a ftop-cock. To fill itwith oxygen from the gazometer, firft place iton the pan full of water, b , and through theftop-cock fuck up the air till the water rifes,and fills it entirely, and then turn the cockwhile the lips ar-e ftill upon it. The jar is thus