i8 



X.4Ti'Kl^ 



[MaV 2, 1895 



musl be iciucmbcreii that il a thcriHoiiieter or a pyrometer, as 

 the case may be, is plunged into a mass of water or of molten 

 metal, the teniperaUire will fall continuously until the water or 

 the metal logins to l>ccome si>lid ; the temperature will then 

 remain constant until the whole mass is solid, when the down- 

 ward course of the temperature is resumed. This little thcrmo- 

 junction is ]>lunj;ed into a mass of gold ; an electric current is. 

 in popular language, generated, and the strength of the current 

 is proportitmal to the temiierature to which the thermo-junction 

 is raised ; s^i that the sikiI of light from a galvanometer to which 

 the ihermo-junction is attached enables us to measure the 

 temperature, or, by the aid of photography, to record any 

 thermal changes that may occur in a heated mas,s of metal or 

 alloy. 



It is only necessary for our purixxse to use a |x>rlion of the long 

 scale, and to make that jiortion of the scale movable. Let me 

 tr\' to trace l)efiire yoii the curve of the freezing of pure gold. It 

 Mill be necessiir)' to mark the |xisition occupied by the movable 

 sjxit of light at regular intervals of time during which the gold is 

 ne>ar 1045', that is. while the metal is Incoming .solid. Everj- 

 time a metronome Iwats a second, the white screen A (Fig. 3). a 

 sheet of |Xi|K-r will be raised a definite number of inches by the 

 gearing and handle, B, and the |Xisilion .succes-sivcly occupied by 

 the spot of light, t'. will lie marked by hand. 



Vou see that the time-tenii>erature curve, -v, j', so traced is not 

 continuous. The freezing |Miint i>f the metal is very clearly 

 marked by the horizontal |mrtion. If the gold is very pure the 

 angles are sharp, if it is impure they are rounded. If the 

 metal had fallen liclow its freezing (xiint without actually be- 

 coming solid, that is. if su]>erfusion or surfiision had occurred, 

 then there would Ix-, as is often the case, a dip where the freezing 

 lx.-gins, and then the teui|K.-ralurc curve rises suddenly. 



If the metal is alloyed with large quantities of other metals, 

 then there nia\ lie several of these freezing ix)ints. as sHccessi\e 

 groups of alloys fall out of solution. The rough diagrammatic 

 method is not sufficiently delicate to enable nie to trace the 

 sulH>rdinate jxiints, bit iheyareof vital importance to the strength 

 of the metal or alloy, and photography enables us to detect them 

 readily. 



Take the case of the tin-cop|x;r series ; you will see that as 

 a ma-ss of tin-cop|K-r alloy cools, there are at least t«o distinct 

 freezing |x)inls. .\l the ujiix-r one the main ma.ss of the fluid alloy 

 liecime solid ; at the lower, sonie definite group of tin and copix-r 

 atoms fall out. the ]K>sition of the lower jioint deiK-nding upon 

 I he com))osilion of the mas.s. 



' ( Ti> f'l- continued. ) 



////■: /.xsTjri'noN OF mechanical 



ENGINEERS. 



""PIIK ordinary spring meeting of the Institulion of Mechanical 

 •^ Kngineers was hehl on Wednesday and Friday evening of 

 litst week, .Viiril 24 and 26. the I'resident, I'rof .\lexander H. 

 W. Kenneily, F. K.S.. occupying the chair Imth evenings. The 

 following was down on the agenda of the meeting : .\djourned 

 rli.scussion on Captain II. Kiall .Sankey's paper on " governing 

 iif Steam Fngines by Throttling and by Variable Kx|>aii.sion " : 

 Ihe •• Third l<e|Hirt to the .Alloys Research Committee," by Prof 

 \V. C. kolxris-.Vuslen. C.H.. F.R.S., " Appendix on the Klim- 

 in.ttion of Impurities during the I'rocess of making ' Kest Selected ' I 

 Copper," by .Mr. Allan Cibb ; " y\p|xndix on the I'yromctric j 

 Kxamination of the .\lloys of Cop|K-r anil Tin." by Mr. .\lfred 

 .•^lansfield. | 



In the discus-sion on Caplain .Sankey's |)a|x-r a nundxT of ' 

 members sinike. A.s a general result it may lie statetl that the 

 (xisilion taken by Ihe author in his |ia|x-r was .sup|xirled. viz. : . 

 that for certain pur|xises. governing by means of the throttle , 

 valve w-i-s to Ix- preferred ; whilst under other conditionsvariable 

 ex|xin.sion guvernors wtiuld have advantages over the other 

 method. Giplain Sankey in hisconlriliulion nn|iartially discussed 

 Uith systems, and his ikiixt n)ay Ix- taken as a good model of 

 what a n)emoir of the knid sliotdd lie, no uruhie bias Ix-ing shown 

 on cither side. ' 



The reiort of I'rof. Kolx-rts-Atistcn w.ts |KTha|is of even 

 greater interest than Ihost which have preceded il': whil.sl the 

 two apixndiic'i of Mes.srs. Ciilili and .Stansfield discussed 

 important practical details. A retpiest had iK-en made that the . 

 inveitigalion* of Warburg ami Tegelmeicr on molecular |xirosily, ' 



Nf>. 



and their oKscrvalions on the '" IClcclroly.sisof Cdass"' should be 

 repeated. It will lie remembered that atoms of .sodium were 

 made to p:vss through glass at a lem]xr,iture of 200 C under the 

 influence of the electric current. Lithium atoms were iheit made 

 to follow along the tracks or molecular galleries left by the- 

 .sodium, the lithium having a lower atomic volume and weight 

 than the sodimn. When jxitassium, having a higher atomic 

 weight and volume, was substituted, it was not foimd (Xissible 

 to trace out the sodium. We are thus, the author said, 

 confronted with a molecular jxirosily which can in a sense- 

 lie gauged, and the meolianical influence of the volume of 

 the atom is thus made evident. It will also be evident 

 that there is a direct connection between the properties of a 

 nuiss and the volume of its atoms. The results previously 

 obtained were entirely confirmetl and somewhat extended in the 

 ex|x;riments the author had undertaken. The septa, or dividing 

 |jartitif>ns. in these fresh experiments, were maile nu^stly of 

 soda glass, of which thick bulbs were blown from barometer' 

 ttdie. In mtist of the experiments the glass waselectrolysed, using 

 mercur}" autl an anutlgam of some metal as cathode and anode- 

 respectively. The temperature was from 250 to 350" C. The 

 electromotive force employed was lOO volts, and the current in 

 the case of the s<Klium exjx'riments averageil about one-lhfiusandth 

 of an ampere, and was sometimes as high as one-fiftieth of ai» 

 aiupere. AVhen the gla.ss Indbs were employed they soon iiecamt 

 cracked, and the free i«.s,s,age of the current fused the gla.ss. 

 forming a well-roimded hole. In each ex|ierinient a safet) 

 fuse was placed in series, to stop the current in case ot 

 break.age. In experiments in which sodium amalgam had 

 been placed in the bulb and pure mercury outside. sodiun\ 

 passed into the mercury to the extent of 0*03 gramme of 

 0'46 grain. In one experiment, which lasted eighteen hours, 

 the amount of sodium found in ihe mercury w.is o'oiji gramme, 

 or 0"2022 grain. The ([uanlily of electricity wliicli passed 

 through the glass was measured by the aid of an electrolytit- 

 cell jilaced in .series, in which co|iper was depositeil to thi- 

 amount of o'02o6 gramme, or o'ji/q grain. Calculating thi- 

 number of cmdombs of electricity passed by means of the 

 electrolysis of gla.ss. the number 55 is found, and by the electro- 

 lysis of cojiper .sulphate, 62 ; thus showing, as well as a rouglv 

 approximate experiment could, that the |)assage of sodium int< • 

 Ihe mercury follows the ordinary law of electrolysis. Il is 

 <loubtfid whether the soilium from the amalgam actually pene 

 trateil right through the glass ; but Ihere can be no ipieslitir. 

 ihat il replaced a consiilerable proportion of the sodium which 

 the glass contained. .\n allempt to |)a.ss potassinm Ihrough 

 ihe .same glass faile<l. Cold was then used, both in the form of 

 amalgam and dissolved in nielallic lead, 1 ml in the latter ca.si. 

 Ihe temperature employeil was, of course, higher. No goldl 

 w.as found to have been transmitted through the glass ; but the 

 glass employed became coloured by gold, and minute s|mngle.s 

 of the metal were found embedded in it. The .same result was 

 obtained when copper was used as an amalgam ; anil in this 

 case minute nodules of copper were dcposiled below Ihe surface 

 of Ihe gla.ss, an efl'cct which is highly suggesllve in connection, 

 with Ihe formation of mineral veins by earth currents. .Sodium 

 amalgam placed in a bulb and surrounded with mercury, but 

 with no current, gave negative results, showing thai simple 

 diflusion did not play any imporlani pari in the results obtained. 

 The fad ihal a current passes al all through glass is a proof thai 

 electrolytic action has taken place: so thai, even if a metal bi 

 luil actually Iransmitleil ihrnngh glass, the )iassagc of a current 

 indicates thai sodium, potassium, or other melallic conslilucnl 

 of the glass, nuisl be leaving il, and is probably replaced by oni 

 or more of the metals in ihe metallic bath which conslilules the 

 anode. 



The author next referred to an .addition made In the recording 

 jiyrometer by means of which increased sensilivene.vs was ob- 

 tained. The galvanonuur. which atVords the means of nuasur- 

 ing the temperatures of ihe masses of metal or alloy undei 

 examination, may occupy one of two positions: il may eilhei 

 be nearer to the slit ihrough which the ray of light falls upon, 

 ihe photographic plate, or it may be further away from it. Il 

 will be evident lhal two galvanometers may be used .simul- 

 laneoiisly, wilh Ihe light from their respective mirrors playing 



1 K. WnrliurH. " Uclur dii; Kkl,lrol>>i: des fcslcil I ll.iscs." WUtlemaiw .' 

 Xnnattn, vol. xxi. 1884. [i. (ni. K. UarlmrK ami V. TiKinin-icr, " llelx'i 

 die clcklrolytinclii: IaIiuiik iks HirKkryslalls," lyiriirniann's AnnaUn, vol. 

 \li.. 189a, pailciS. K. \V:irliuri;. " VelxTiim; Mclhodc N.ltriuin Mclall in. 

 ((ck^lcnchc Kfllircn lin/ufillircii. " IVinlriiiaiin s tiiiittleii. vol. vl. i3(.o. 

 [i.ilif 1. 





