jg Mr. E. Matthey .
Further Contribution s \#163 ; Jan. 8 , ; ( 2 ) ' ' The boundary conditions are : continuity of magnetic induction and electric displacement perpendicular to the reflecting surface , the latter of which is equivalent to continuity of magnetic force perpendicular to the plane of incidence ; continuity of magnetic force along the line of intersection of the plane of incidence with the reflecting surface ; continuity of the rate at which energy flows across the reflecting surface .
Now the refracted light consists of two waves , circularly polarised in opposite directions , and the reflected light is elliptically polarised ; we have , therefore , four equations to determine the amplitudes of the two refracted waves , and the amplitudes of the two components of the reflected wave .
The results of the paper agree with Dr. Kerr 's experiments in the following particulars : \#151 ; ( i. ) The reflected light is elliptically polarised .
( ii .
) When the magnetisation is parallel to the reflecting surface , no effect is produced when the incidence is normal , or when the plane of incidence is perpendicular to the direction of magnetisation .
( iii .
) When the plane of incidence is parallel to the direction of magnetisation , and the light is polarised in the plane of incidence , the magnetic term increases from grazing incidence to a maximum value , and then decreases to normal incidence .
The principal point of disagreement is , that in all cases the intensity of the reflected light is unchanged when the direction of the magnetising current is reversed .
I do not think that the results of the theory can be considered altogether unsatisfactory , since they certainly explain some of Dr. Kerr 's experimental results ; and I am disposed to think that the disagreement is due to the disturbing influence of metallic reflection .
At the same time , the question is one which can only be decided by experiment , and it is therefore most desirable that experiments on magnetised solutions should be made .
IV .
" Further Contributions to the Metallurgy of Bismuth .
" By Edward Matthey , F.S.A. , F.C.S. , Assoc. Roy .
Sch .
Mines .
Communicated by Sir G. G. Stokes , Bart. , F.R.S. Received December 22 , 1890 .
In October , 1887 , I read a paper before the Royal Society* upon a new method which I incidentally discovered while working with a view to separate copper from bismuth , by fusion with bismuth sulphide .
* * Roy .
Soc. Proc. , ' vol. 43 , p. 172 .
1891 .
] to the Metallurgy of Bismuth .
79 I stated in this paper that bismuth " frequently contains a small proportion of copper , an element most detrimental even in small traces , and hitherto only eliminated by a wet process , costly in practice and tedious in operation .
It is necessary by such method to dissolve up the whole of the alloy , and precipitate the bismuth in the usual manner\#151 ; a bulky operation , and one requiring a considerable amount of time .
It became therefore advisable , in order to treat cupriferous bismuth rapidly and upon a commercial scale , to effect this separation , if possible , by means of a dry process .
" In further researches in the metallurgy of this interesting metal , a case was found in which bismuth contained a very small proportion of copper , under 0'5 per cent. , but sufficient to render the metal useless , and in fact , to destroy those characteristic properties upon which its industrial applications depend .
Instead of treating this cupriferous bismuth by fusion with bismuth sulphide , which necessitates a temperature sufficiently elevated to bring about a complete fusion of the bismuth sulphide , and consequently , unless very great care be taken , great loss by volatilisation of bismuth , it occurred to me to fuse the alloy , and , at a temperature a little above its melting point , to add a small proportion of sodium monosulphide .
The mass was then stirred well , so as to bring every portion of the fused alloy into contact with the fused sulphide .
After about one hour 's stirring , a test was made of the molten metal , and it was found that the amount of copper in it wras very considerably decreased .
By skimming off the film of scoria which had risen to the surface , adding a further small proportion of the sodium monosulphide , and continuing the operation of stirring , every trace of copper was eliminated , and the bismuth so freed from copper rendered in every way suitable for industrial use .
The first experiment was made upon a quantity of 105 kilograms , which yielded 94 kilograms of bismuth free from copper , and about 11 kilograms of skimmings containing the whole of the copper , their bismuth contents of course being available for reduction with further and larger quantities of skimmings as they accumulated .
Ibis process has been repeated upon very considerable quantities j of cupriferous bismuth , and has proved to be successful .
This question of keeping the temperature low is of much import-j \amp ; nce , for the lower the temperature the less tendency there is for the bismuth to volatilise , and as it is necessary to obtain the bismuth j fxee from traces of impurity , which entirely change its nature , it will be seen that any improvement in manipulation , or in the process i itself , which enables pure metal to be obtained possesses much interest .

