•36 Apparatus for Experiments -with the Blow-Pipe. 



for extruding the air will anfwer to about a quarter of a pound avoirdupois upon each 

 fquare inch of furface, which is not more than an eighth part of the preflure in the regu- 

 lating belly o f tlie blowing machines at our great foundries. I think the quantity of one 

 cubic inch per fecond is quite as much and probably more than iflues out of the blow-pipe 

 in any courfe of experiment. A pair of bellows capable of extruding fomewhat more than 

 two cubic incites at a flroke, would confequently fupplyas much air as the pipe would de- 

 liver, provided the ftrokes fuccecded each other about once in two feconds, which appears 

 to be a convenient rate of working, and by no means too quick. In very fmall bellows the 

 internal contents may be eflimated at one third of the contents of a parallelopipedon, or 

 fquare box capable of circunifcribing the bellows when open. As a full allowance, let 

 us fuppofe the contents of this imaginary box to be eight cubic inches, and its depth two 

 inches. Its upper furface muft then be equal to four fuperficial inches. That is to fay, a 

 "pair of bellows fully fufficient to fupply the blow-pipe will not require larger dimenfions 

 than three inches in length, one inch and a half in width, with a lift of one inch and a half 

 for each flroke. The fame efFeiSl may be produced by a fyringe one inch in diameter, and 

 two inches and a half long. 



The refervoir for the air mp.y either be another pair of bellows rather larger, and difpofed 

 to fhut by a weight or a fpring; or it may be fimply a veflel of fufficient capacity to 

 receive the air from the bellows, and emit i^ in a conftant ftream by virtue of its fpring 

 under condenfation, in the fame manner as water is emitted from the air veflel of a fire- 

 engine. TliC emifTion of air from fuch a veflel may, in a loofe way, be taken to be at half 

 the velocity of its introdudlion ; and, confequently, at the end of every ftroke the veflel 

 will contain half as much more air, than the medium ftate of condenfation, as amounts to 

 the whole contents of the bellows ; and at the beginning of every ftroke, the veflTel will 

 contain half as much lefs than that quantity. From thefe confiderations, it appears that 

 the fteady ftream from a fire-engine could not be produced but by two pumps alternately 

 acling, and that the air anfwers fcarcely any other purpofe in the fmall air-veflels of thofe 

 engines, than that of rendering the intFufion of the non-elaftic water lefs fudden and "vio- 

 lent. It will alfo appear from calculation, (as it does from prailice, in the regulating bellies 

 of blowing engines, on which the re-aftion is afforded by a variable force fomewhat refem- 

 bling the fpring of tlie air) that a very large veflel is required to give a practical uniformity 

 to the emitted ftream, when the aiflion of intrufion is not conftant. Hence it appears, 

 that an air vtfl'el cannot be ufcd to advantage, unlefs the bellows to our fmall apparatus be 

 made double by two fixed outer boards, with valves opening inwards, and a moveable 

 diaphragm alternating between them, and forcing the air through valves opening into one 

 common nozle— or unlefs the fyringe, if ufed, be made to adl both ways, namely, by the 

 returning as well as by the direct ftroke. A very neat and compadl; apparatus for the 

 blow -pipe might be made in this way; but it is probable that the confiderations and their 

 advantages may be more particularly applicable to the large engines at the fmelting- 

 works. 



XL AJ}.ort 



