70 Royal Society. 



another paper ; namely, tliat there are slow contibustions with- 

 out flame, and which take place at temperatures below the point 

 of the visible ignition of metal. He determines in this part of the 

 paper, that the electrical spark or flame does not produce heat 

 by com])ression, but by expansion by an acttial increase of tem- 

 perature. 



In the third section he shows that different elastic fluids have 

 different effects in extinguishing flame ; that nitrous oxide is the 

 lowest, olefiant gas the highe^t, as to this power : and this does 

 not depend upon capacitv for heat or density ; but on an actual 

 power of abstracting heat, which is much highest in the combus- 

 tible gases, and which seems analogous to conduciing power in 

 solids and fluids. Steam has very low powers of preventing ex- 

 Jilosion, and azote has low powers compared to inflammable 

 gases. The increased cooling power of the azote in condensed 

 niixtufes prevents the comlnistion from increasing very rapidly,, 

 and the diminished cooling power in rarefied atmospheres inter- 

 feres wiih a rapid diminution of the heat of combustion ; so that 

 at all pressures which can occur at the surface of the earth, the 

 atmosphere still retains the same relations to combustion. 



Sir Humphry began the fourth section by illustrating the sim- 

 ple explanation which he has always given of the operation of wire 

 gauze in intercepting flame, by some decisive experiments. He 

 pointed out the limits of the size of the wire and the mesh, which 

 differ for different flames; — that of phosphuretted hydrogen and 

 sulphur requiring the smallest mesh, and that of carburetted hy- 

 drogen or fire damp being stopped comparatively by the largest 

 mesh; and the law being always as the heal required for burning 

 the gas, and as the heat produced by the combustion of the gas. 

 To prevent explosion from passing trom a mixture of oxygen and 

 hydrogen requires 7^*00 or SOOO apertures to the square inch, 

 whilst carburetted hydrogen and atmospheric air will not pass 

 their explosion through a mesh containing 700 or 800. He re- 

 fers to the form of the miners wire-gauze lamp, which he has 

 adopted after many experiments, and which under all circum- 

 stances has been found to offer absolute security. By the in- 

 crease of the combustibility of gaseous mixtures hy heat, he ex- 

 plained the effects of various lamps and fire-places. 



He stated, i:: spealting of meteoric appearances, that falling 

 stars could not be owing to the combustion o{' gaseous meteors; 

 but that they must be solid ignited masses moving with great ve- 

 locity in the upper regions of the atmosphere. — It would exceed 

 our limits to attempt to enter into all the apphcations of these 

 researches ; and indeed such an attempt would be incompatible 

 with the nature of the investigations, many of which were too 



profound 



i 



