174 ANNUAL OF SCIENTIFIC DISCOVEKY. 



(see his " Treatise on the ^rechanieal Tlioory of Heat"). The 

 mathematical law of the expansion in this case can be given with 

 perfect prcicision ; but its ciivunisiances are not accurately realized 

 in practice, because the cylinder is always made of a raiJidly-con- 

 ductin<;^ material. 



" 5. When the steam expands and performs work in a conduct- 

 injj cylinder, which receives no supply of heat from without, l)ut 

 is left to undergo a great alternate rise and fall of temperature 

 through its alternate connection with the boihu' and the con- 

 denser, the law of expansion becomes ver}' variable, and the 

 prol)lem of determining it extremely complex. It is certain, how-' 

 ^ver, that a grt'at waste of heat occurs in every case of this kind, 

 as Mr. Isherwood's experiments liave shown. In a pa23er read 

 to tiio Institution of Engineers in Scotland, about two years ago, 

 I discussed some of Mr. Isherwood's earlier experiments, and 

 showed that tiiey gave jjroof of a waste of heat increasing with 

 the fall of temperature due to the expansion of the steam, with 

 the extent of conducting surface of the cylinder, and with the 

 duration of the contact between the hot boiler steam and that 

 conducting sui'face." 



NEW DETECTOR OF FIRE-DAMP. 



Mr. G. F. Ansell, of the Royal Mint, has proposed a novel 

 application of Professor Graham's law of gas diifusion for the pur- 

 pose of ascertaining and giving warning of the presence of accu- 

 mulations of lire-damp in coal mines. The apparatus described 

 by Mr. Ansell is a glass U tube, having one aperture closed with 

 a plate of graphite, or equivalent porous diaphragm, and a few 

 inches of mercury in the bend. If such an arrangement, filled 

 in the first instance with air, be placed under the influence of an 

 atmosphere containing five per cent., or even less, of light carbu- 

 retted hydrogen or marsh gas, the presence of such admixture 

 will be instantly detected by the passage of the gas thi'ough the 

 interstices of the gi'aphite, and the consequent expansion in vol- 

 ume of the gaseous contents in the tube ; the column of mercury 

 then rises in the oi^posite limb of the apjiaratus, and is made to 

 record itself either by completing the circuit of a voltaic alarum, 

 by deflecting a galvanometer needle, or, lastly, by the adaptation 

 of the simpler mechanism of a wheel barometer. We understand 

 that the invention has been patented by Mr. Ansell, and, inasmuch 

 as it gives great promise of successful employment, the apparatus 

 must be deemed AveU worthy of immediate trial. — Reader. 



CONTROLLING CLOCK. 



Mr. Lang, C.E., in a paper read before the Royal Scottish Soci- 

 ety of Arts, proposes a new method of adjusting a clock to within 

 a small fraction of a second, which may perhaps be found useful 

 in the case of a controlling clock, where it is important not to 

 throw the controlled clocks out of time. It consists essentially in 

 varying the virtual length of the pendulum by a very small 



