On Surfusion in Metals and Alloys .
'On Surfusion in Metals and Alloys .
" By W.C. Roberts-Austen , C.B. , D.C.L. , F.R.S. Received May 20 , \#151 ; Read May 26 , 1898 .
( Plate3 8 , 9 .
) Introduction .
The fact that metals and alloys may be maintained in a fluid state at temperatures which are many degrees below their true freezing points , has been but little studied .
As regards salts the question of surfusion has recently received much attention .
Ostwald* has shown , as the result of an investigation of great interest , that a very minute quantity of a solid will cause a mass of the same substance to pass from the surfused to the solid state .
His work , moreover , has led him to distinguish between the meta-stable , or ordinary condition in which surfusion takes place , and the labile condition which occurs at temperatures much below the melting point .
Ostwald 's paper , and one by M. Brillouin , t on the theory of complete and pasty fusion lead me to offer the Royal Society the results of some experiments of my own on the surfusion of metals .
Historical .
) Metals do not appear to have been studied from the point of view of surfusion until the year 1880 , when some excellent experiments ; on the surfusion of gold were made by the late Dr. A. D. van i Riemsdijk , J by whose early death , which occurred last year , Holland has lost a skilful physicist .
He pointed out that:\#151 ; [ r " Faraday , in his memoir on regelation , published in 1858 , stated that acetic acid , sulphur , phosphorus , many metals and many solu-j tions , may be cooled below the congealing temperature prior to solidifi-j cation of the first portions."S On the other hand , in their treatises ' on physics , Danguin|| and Jamin^f mention tin as the only metal ij which is capable of remaining liquid at a temperature 2 5 below !
the true melting point of the metal , which is 228 ' C. Van Riemsdijk 's contribution to the subject of surfusion of inetals ) consisted in showing that the well-known phenomenon of eclair , the I brilliant flash of light which often attends the solidification of the i metal in the ordinary assay of gold , is really due to surfusion .
He * * Zeit .
fur Physikal .
Chem. , ' 1897 , vol. 22 , p. 3 .
f 'Ann .
de Chim .
et de Phys. , * 1898 , vol. 13 , p. 264 .
X * Ann. de Chim .
et de Phys. , ' 1880 , vol. 20 , p. 66 .
S * Experimental Researches in Chemistry and Physics , ' p. 379 .
|| Vol. 1 , 1855 , p. 892 .
% Vol. 1 , 1859 , p. 105 .
Dr. W. C. Roberts-Austen .
also pointed out that surfusion could be either stimulated or hindere ; by suitably modifying the conditions , but he made no attempt a thermal measurements .
It was not until ten years after the publica tion of van Riemsdijk 's work that the recording pyrometer , which ] devised and submitted to the Royal Society in 1891 , * enabled sucl measurements to be readily effected .
It consists of a camera , enclosing a dead-beat galvanometer , t\lt ; which the free ends of a thermo-junction are attached .
The thermo junction is suitably protected , and , as it only consists of two wires twisted together and covered with a fine clay tube , it can be placed in the cooling mass of molten metal or alloy , and the cooling curvt of the mass may be traced by a spot of light from the galvanometer which falls on a moving sensitised plate .
A ready method foi studying all the phenomena of the solidification of metals and alloys is thus afforded .
The freezing point of a metal , or the initial freezing point of an alloy , for a fluid mass of two metals may possess many points ol solidification , is represented by one or other of three typical curves .
These are shown in the accompanying figures which indicate the Fig. 1 .
nature of the curves , traced by the recording pyrometer .
Fig 1 shows the freezing point curve of a pure metal , the horizontal portion , ab , indicating the actual solidification of the mass , the sharpness of the angles at a and b attesting the purity of the metal .
The initial freezing point of most alloys would resemble fig. 1 in having the corner a sharp , while the point b is generally rounded off .
If the alloyed metals form an isomorphous mixture , neither angle is sharp , and in many cases there is no true freezing point , the curve being of the form shown in fig. 2 .
This represents the freezing of the gold-silver alloy containing 28 per cent , of gold in which the fluid mass , as a whole , passes through a long pasty range ' Proc. Roy .
Soc. , ' 1891 , vol. 49 , p. 347 .
On Surfusion in Metals and Alloys .
Fig. 2 .
before solidifying .
The third type of curve , which may be a modification of the other two types , indicates the occurrence of surfusion , and , as a case in point , pure gold has been taken , the bend at a , fig. 3 , showing the amount of surfusion which was observed .
I have detected pronounced cases of surfusion not only in gold but in copper , bismuth , antimony , lead , and tin .
Surfusion , moreover , is not confined to pure metals , and I showed in 1893* that the eutectic alloy in the bismuth-copper series presents a marked case of surfusion .
Fia .
3 .
-t pi rn \gt ; o : Pi TIME , In order to study surfusion , it is necessary to make the galvanometer , to which the thermo-juncfcion is attached , very sensitive , and , by suitable adjustment , it is easy to catch on the sensitised plate any desired portion of the range of the spot of light .
It is , however , preferable to balance by a potentiometer the current which results from the heating of the thermo-junction , and in this way to prevent the mirror from swinging through a long range .
The sensitiveness of the instrument is but slightly diminished by the introduction of the potentiometer .
As the thermo-junction cools down , the spot of light from the galvanometer is simply made to traverse a short distance many times , instead of a long range once .
The sensitised * " Second Report , Alloys Research Committee , " * Mech. Eng. , ' 1893 , Plate 32 .
Dr. W. C. Roberts-Austen .
plate need only be exposed to the action of the spot of light at the critical moment , when solidification or surfusion is known to bt imminent .
A paper by my assistant , Mr. Stansfield , illustrating the use of this method , will shortly be published by the Physica Society .
A curve , traced by the aid of either of the sensitive methods which have just been described , if it represents the surfusion of t metal or alloy does not merely show a slight depression as in the case of pure gold shown at a , fig. 3 : the slight depression becomes e deep dip , ( Plate 8 , fig. 1 , which represents the surfusion of gold ) In this case three curves were taken on one plate , the line al represents the heating and melting of gold , the horizontal portion marking the actual melting of the metal .
The lines and fgh .
on the other hand , represent successive coolings and solidifications of the metal , surfusion occurring in both cases , the horizontal portions at e and h representing the beginning of solidification of the gold in each case .
It is noteworthy that these successive points of solidification differ by less than half a degree , the melting point of the metal ( the line aV ) occupying a median position between them .
During the surfusion , the temperature of the metal fell about 2 ' below its true freezing point .
Pigs .
2 and 3 , Plate 8 , both represent the surfusion of gold ; fig. 4 , Plate 8 , that of copper* while fig. 5 , Plate 8 , represents the surfusion of an alloy of antimony with 25 per cent , of copper which freezes at 520 ' .
It was selected as representing the highest temperature of surfusion which I have as yet examined in the case of an alloy .
Such a surfusion curve , in the case of an alloy , may do much more than reveal the sudden release of the latent heat of the fluid mass .
The curve may present a complicated record of modifications in molecular grouping occurring in the brief space of time actually occupied by surfusion .
This is shown by the points d , e , in fig. 6 , Plate 9 , which represents the surfusion of tin alloyed with 36-5 per cent , of lead .
All the curves on Plate 9 serve as bonds connecting the behaviour of alloys with that of freezing solutions of salts .
The explanation of the existence of these points , d , e , is simple .
The freezing point curve of the lead-tin series is a very ordinary type , and is shown in the accompanying diagram ( fig. 4 ) , and , as regards the portion where the lines meet , full details are given on a larger scale in fig. 5 .
It consists of two branches united at the point where they join a horizontal line which represents the freezing of the * Figs. 2 , 3 , 4 on Plate 8 were taken with an insensitive galvanometer and have been enlarged by photography from the original plates .
Fig. 5 , Plate 8 , and all the eurves on Plate 9 are given exactly as they were taken on the plate of the recording instrument .
The co-ordinates , as in the case of the figures 1 to 4 , are time and temperature .
On Surfusion in Metals and Alloys .
Fig. 4 .
eutectic alloy of the lead-tin series .
This alloy freezes at a constant temperature ( 183 ' ) .
The eutectic constitutes a fluid residual " liquor " which is left after the excess of either lead or tin has fallen out as the mass cools down .
When , however , the tin is present in slight excess of the amount required to constitute the eutectic , the whole mass of the alloy may remain fluid at temperatures below that at which the eutectic would naturally freeze .
It may even cool to a temperature at which it can no longer maintain all its lead in solution , and some lead will , therefore , fall out while surfusion is actually taking place .
This explains the existence of the point d , which marks an arrest in the fall of temperature , in Plate 9 , fig. 6 ( on the line ab , representing the fall of temperature during surfasion ) .
Conversely , when the surfusion is ended , the latent heat is released , and the line bee represents the rise of temperature due to the release of latent heat .
There should also he on this line an indication of a retardation or arrest in the rise of temperature , because the lead which fell out of solution ( at the point d ) has to be rS^melted .
This point of larrest will be found at e. In the same way , in fig. 7 , Plate 9 , which represents a fine case of surfusion in the lead-tin alloy containing 64 per cent , of tin , there is also a point at / , on the line gh , at which point the lead fell out of solution during the surfusion .
Fig. 8 , Plate 9 , represents the surfusion of a lead-tin alloy containing 68 per cent , of tin .
In this case the tin is in large excess , and freezes first .
In becoming solid , the tin would have been represented by a horizontal line somewhere about the point s , on the line ij , had not surfusion occurred , and been terminated by the solidification of tin at Tc .
This is followed by the solidification of the eutectic mother liquor represented by the line Im. This eutectic does not surfuse because the tin ( the metal in this alloy which is prone to surfuse ) is already present in the crystalline form .
In fig. 9 , Plate 9 , a similar case is represented , but a small , though | distinct , surfusion at the point n has been recorded .
This surfusion Dr. W. C. Roberts-Austen .
at n probably indicates that the lead in the eutectic may surfm slightly even though the crystallisation of tin has begun .
By taking a series of lead-tin alloys which do not contain more tha 2 or 3 per cent , of tin , in addition to the amount required to constitul the eutectic ( 62 per cent , of tin ) , the temperatures at which th lead begins to crystallise below 183 ' , can be measured .
The result are shown in the diagram , fig. 5 .
In it the co-ordinates are ten Fig. 5.\#151 ; Freezing-point curves of lead-tin alloys .
Tin L . .
\#166 ; ( \gt ; --- q\gt ; i \#166 ; \#166 ; 1 Percentage of peratures and percentages of tin .
Each alloy examined has , it wil be seen , at least two freezing points , and some appear to have three but in all the alloys one of the freezing points ( the eutectic ) is ai the constant temperature 183 ' .
In the eutectic alloy ( 62 per cent , o tin ) , theoretically these points , d , e , should coincide .
In fig. 5 thei are not quite coincident , and this is due to the fact tha.t tht lines AB , CD , drawn through the observed freezing points , are o !
necessity somewhat lower than the ideal solubility curves .
Tht freezing points which lie below the horizontal or eutectic line wert recorded while surfusion occurred .
For instance , take the alloj containing 64 per cent , of tin , its initial freezing point would be ai a ( 185 ' ) , that of its eutectic at b ( 183 ' ) , while the third freezing point , c ( 176 ' ) , is due , as has already been explained , to the falling out of lead while the fluid mass was in the surfused condition .
In order to observe the uppermost of the three points ( 185 ' ) , it is necessary to stir the mass to prevent the possibility of surfusion .
On the other hand , a record of the lowest freezing point can only be obtained in a fluid mass which is perfectly tranquil .
When lines are drawn through all the points recorded it will be seen that the line AB , representing the solidification of lead , extends far below the horizontal line XX .
It should be remembered that if the fluid mass SURFUSION .
Rcy Soc. Proc. VPlate , Fig.2 .
Gold .
Fig. 3 .
Gold .
Fig. 4 .
Copper .
Fig. 5 .
Antimony with 25/ Copper .
SIR FUSION .
Roy Soc. Proc. Voi .
63 , Plate 9 .
^ Austen On Surfusion in Metals and A lions .
45 \amp ; 3 stirred , the first point to be recorded will be the upper or tin point 185 ' ) , followed by the eutectic point ( 183 ' ) .
If the alloy surfuses , he first point recorded will be the lowest point ( 176 ' ) , followed in urn by the eutectic .
In general , therefore , only two freezing points an be obtained in a single record .
In the case of salts the crossing of the curves of solubility has dready been observed by H. le Chatelier and by Dalims , but in the as of alloys , experimental evidence has hitherto been wanting .
Che silver-copper series presents many analogies to the lead-tin eries .
Hey cock and Neville* , in their excellent work on the com-jlete freezing point curves of many series of alloys , have calculated ivhat the ideal freezing point curve of the silver-copper series would :\gt ; e , but the present paper affords , I believe , the first experimental widence as to the identity of the behaviour of saline solutions and netallic alloys as regards selective snrfusion .
Prolonged experience in these methods of manipulation may prove hat it is possible to effect the separation of a particular metal or lefinite groups of metals by dropping in ( during the surfusion of the luid mass ) a fragment of the same metal or of the particular group \gt ; f the associated metals it is intended to separate .
It is well known hat the introduction of a fragment of the same metal or of an iso-norphous metal or alloy Avill determine its solidification .
Such a nethod may readily be employed in studying the surfusion of salts .
In the case of metals , so far as my own experiments go , the .
sur-rused state is singularly unstable , for it may be disturbed even by srery slight tremors .
It remains to be seen whether it is possible to arrange the experiments in such a way as to maintain metals and alloys for an indefinite time in Ostwald 's meta-stable condition which would need the presence of a particle of solid matter to induce the solidification of the mass .
Ostwald applies to the change from liquid to solid the equation which represents the gas-liquid change .
In the former case there are , however , the three phases , solid , liquid and gas , present , and a complete expression of the change must take account of all three .
Thus during surfusion the gas phase is in equilibrium with the liquid phase , but when solidification has begun the gas phase must also be in equilibrium with the solid phase .
It is possible that the solidification of the mass may be started by crystals deposited directly from its vapour so that solidification of a surfused metal may be started by crystals from its own metallic atmosphere .
The amount of gas evolved by a solid metal is , of course , very small .
Evidence of the vaporization of metals at very moderate temperatures is not wanting .
Demar^yj showed in 1882 that vacuo metals evaporate * 'Phil .
Trans. , ' A , vol. 189 ( 1897 ) , p. 25 .
+ c Comptes Eendus/ 1882 , vol. 95 , p. 183 .
Prof. D. Mendeleeff .
Experimental Fio .
6 .
at much lower temperatures than they do at ordinary atmospher pressure .
Pellat* pointed out that the proximity of the surfat of a metal to that of another metal in air , changes its electrical coi dition , and he attributed this to vaporization of metals , showir that even iron exerted an influence at a distance .
Colsonf show that a photographic plate was affected by the vapour of certai metals , notably by zinc , cadmium , and magnesium even throng porous septa .
Dr. Russell } : in some recent and very interests experiments , was led to the conclusion that even infusible a metal as cobalt will by vaporization , affe\lt ; a photographic plate .
In January , 1897 , before hearin of Dr. Russell 's experiments , I began some exper ments with a view to ascertain whether metals , vapoi ized in vacuo near the ordinary temperature , wi actually unite to form alloys .
The arrangement i shown in fig. 6 .
A and D are two discs of meta with polished surfaces separated by a ring of glass , th whole being enclosed in a vacuous tube which could b heated in a water bath .
I found that when cadmiun and silver were opposed for eight days at a tempera ture of 50 ' an appreciable deposit of a tinted cadmium silver alloy formed on the surface of the silvei Cadmium must , therefore , have passed across th interval between the discs A and B. The results given in the present paper , reveal addi tional points of similarity between the behaviour o alloys and that of ordinary saline solutions .
I trust , therefore , tha it may be useful as a continuation of ray investigation on th* Diffusion of Metals , " which formed the subject of the Bakeriai Lecture of 1896 .
MExperimental Investigations on the Oscillations of Balances .
' By D. Mendeli^eff , For .
Mem. R.S. Received June 9 , \#151 ; Read June 9 , 1898 .
In the year 1893 the Central Chamber of Weights and Measures .(Glavnaya Palata Mer y Vesov ) was created in St. Petersburg to act as a Central Institution of the Empire for the verification of all kinds of standard measures .
Having been appointed Director oi the above-mentioned Institution , I was first of all occupied ir * 'Comptes Rend us , ' 1882 , ycI .
94 , p. 1247 ; 1890 , vol. 123 , p. 104 ; and 1898 , vol. 126 , p. 1338 .
t Ibid. , 1896 , vol. 123 , p. 49 .
' t ' Roy .
Soc. Proc. , ' 1897 , vol. 61 , p. 424 , and ibid. , vol. 63 , p. 102 , Bakerian Lecture , 1898 .

