TRANSACTIOXS OP THE SECTIONS. 209 



nearost the fire, and to act upon a piston under full pressure through part of its 

 stroke ; it is then cut off and allowed to expand until the pressure is a little above 

 that of the air, hut it is still much hotter than the air ; it then enters an air-chain- 

 her, and part of it is used for the blast of the fire, which is in a firebrick chamber 

 without bars, the ashes fusing and fiaming ofi" as slag : tlie remainder of the spent 

 air not used as a blast is mixed with the products of combustion immediately above 

 the fire and before they enter the fiue, thus diminishing the intensity of the heat 

 and preventing its injuring the tubes, and also using up the heat of the spent air in 

 helping to heat a fresh supply. 



The complete cycle consists of the following steps : — Air is compressed, is forced 

 into tub 3.?, is then heated, then acts upon a piston, and a part of the spent air, which 

 still retains considerable heat, is used as the blast and the rest in heating a fresh 

 supply ; thus the real work at our disposal is the ditt'erence between the compres- 

 sing and working cylinders. 



The above description applies to an engine where air is used ; when nitrogen is 

 used, the same gas is used over and over again, and slight modifications are neces- 

 sary, but the general principle is the same. 



By tliesa means it is hoped that a considerable amount of the loss of heat that 

 must accompany the use of steam may be avoided, especially that which passes 

 up the chimney and that which passes away with the water that becomes heated 

 in condensing the steam, an amount many times that which is converted into work 

 in the steam-engine. 



Oil a New System of Testinj the Qml'dij of the Malleahh Metals and Alloys, 

 with E.vjterimental Illustrations. By GtJSiAv Ui.scnoF, Jan. 



Superior qualities of malleable metals and alloys are characterized by their being 

 able permanently to extend in all directions by rolling or hammering without rup- 

 ture, whilst inferior qualities break before reaching the maximum of extension 

 which the former can endure. The cause of this is the dilFerence of the cohesive 

 power in different qualities of the same metal or alloy. From this it must be con- 

 cluded that if different qualities of the same metal or alloy have been rolled in 

 exactly the same manner, the better their quality the oftener they can, after roll- 

 iug, hi bent in reverse directions at a certain angle without breaking. Upon this 

 principle my te:^t is based. 



The to-and-fro movements which, for instance, pure zinc, the best and the most 

 inferior quality of commercial zinc hitherto tested, withstood without breaking, 

 under the conditions presently described, wei-e respectively in the ratio of 100, 54, 

 and 19; for a good quality of steel, charcoal iron prepared by puddling, and ordi- 

 nary bar iron the ratio was 100, 43, and 26 ; for ditterent qualities of commercial 

 copper between 100 and 19 ; for different qualities of commercial tin between 100 

 and 10, &c. It is obvious that once having such figures as standards, auj' other 

 figure obtained in testing metals or alloys will indicate their quality in reference to 

 the quality of the standard, the metliod being the more reliable, as properties 

 which are essential for practical purposes form the criterion. 



The accuracy and saving of time and labour through this test will be estimated 

 from the fact, that, for instance, the deteriorating effect of as little as 1 part of tin 

 upon 10 million parts of pure zinc, or of 1 part of cadmium upon 2o0,000 parts of 

 pure zinc, can be detected with certainty in less than one hour, whilst such traces 

 would probabl}'' escape the notice of a chemist if he spent a fortnight in analvzing. 



As the rolling of samples for comparative tests must always be efl'ected "under 

 the same circumstances, the shape of the different samples must be alike before 

 the rolling is commenced. Metals which do not, or not materially, alter in quality 

 through fusion, are melted and poured into iron moulds of the required size, whilst 

 metals whicli alter in quality through fusion are cut into bars of the required size. 

 Tlieso bars are then, with or without annealing, according to the nature of tlie 

 uijtal or alloj', rolled in a test rolling-mill until they have a standard weight. 



To the central screw of the rolling-mill which actuates the two screws which 

 press upon the upper roller a dial-plate with pointer is attached, enablinf one to 



1S70. 14 



