200 On the Dispersion of Artificial Double Refraction .
By L. N. G. Filon , M.A. , D.Sc .
, Fellow and Lecturer in Mathematics of University College , London .
( Communicated by Professor F. T. Trouton , F.R.S. Received January 25 , \#151 ; Read February 28 , 1907 .
) ( Abstract .
) 1 .
It is well known that glass compressed unequally in different directions , behaves like a crystal whose optic axis is along the line of stress .
If Ti , T* are the principal stresses in the wave front , ya2 the refractive indices of the two rays for which the directions of vibration are along Ti , T2 respectively , then the relative retardation of the two oppositely polarised rays is R = ( / xi\#151 ; fi2)r = C(Ti\#151 ; T2)t , where r is the thickness of glass traversed .
C may be called the " stress-optical coefficient " of the glass .
It differs for different glasses and in the same glass for different colours , but it is usually assumed independent of the value of the stress .
2 .
The experiments described in the paper were undertaken with a view to finding , for a number of glasses of different chemical composition , in particular borosilicates , how the coefficient C varied with the colour of the light used .
No direct compression experiment could be employed , because it has been found impossible in such experiments to obtain a distribution of stress which remains uniform for all loads .
A new method was therefore devised which consisted in passing the light through two beams of glass placed at different levels and bent in a vertical plane by bending moments equal in magnitude and opposite in sense .
It may be shown that if the coefficient C is , for monochromatic light , independent of the stress , such a combination is optically equivalent to a single slab under uniform compression or tension , according to the adjustment of level .
Light passed through such a combination between crossed Nicols is then analysed by a spectroscope .
The spectrum is crossed by dark bands corresponding to wave-lengths for which R = integer multiple of X. The measurement of X then gives the required relative retardation .
The description of the method and the investigation of the various corrections fill the first half of the paper .
3 .
The second half is devoted to the experimental results and their On the Dispersion of Artificial Double .
201 discussion .
It is found that the observations are fairly fitted , on the average , by the formula C Co 1-Xo/ X ' where Co , X0 are constants .
This formula cannot be justified theoretically and is to be regarded as an empirical formula , valid only within the limits of the visible spectrum , and then probably not for all glasses .
If , however , we try to fit some simple formuhe which are suggested by theoretical considerations , we find that they do not fit so well .
The true formula is probably of the form C i-(W ' where \p = wave-length , in vacuo , of light belonging to one of the free periods of the atoms of the glass .
Consideration of the constants in one case suggests that the free periods which are predominant in causing ordinary dispersion in the glass are different from those which are predominant in causing the dispersion of artificial double-refraction .
4 .
The method allowed the effects of tension and of pressure to be observed separately . .
The constants Co , X0 were found in each case to be sensibly different for tension and pressure .
C0 is usually greater for tension than for pressure , and Xo is invariably less .
This difference , however , might also be accounted for by differences of adjustment which render the experiments for tension and pressure not strictly comparable .
Until further data are available the effect must remain doubtful .
5 .
The results of careful observation have shown that in at least two glasses systematic deviations exist from the law C \#151 ; C0/ ( l\#151 ; Xo/ X ) .
These deviations are local in character , and their study leads to the conclusion that they are probably due to an absorption band corresponding to a free period efficient in producing double-refraction under stress .
The results seem further to indicate that the effect of the stress is to change the number of the electrons which respond to vibrations in a plane through the line of stress relatively to the number of electrons responding to vibrations perpendicular to the line of stress , the actual periods of vibration being unaltered .
6 .
Analysis of the chemical composition of the glasses examined shows 202 Messrs. C. Cuthbertson and E. P. Metcalfe .
[ Oct. 23 , that B2O3 increases and K20 decreases the coefficient C0 .
\0\gt ; on the other hand , does not seem to vary much for the range of chemical composition of the glasses in question .
For such glasses , then , the dispersion of doublerefraction appears roughly proportional to the stress-optical coefficient .
* 7 .
The experiments have also revealed that in one glass , for which the relation of stress to molar strain departed sensibly from linearity , the optical effect was proportional to the stress\#151 ; that is to the molecular strain\#151 ; and not to the molar strain ; a result in accordance with what one might expect from the phenomena of double-refraction in viscous fluids in motion .
On the Refractive Indices of Gaseous , , Cadmium , Mercury , Arsenic , Seleniu , and Tellurium .
By C. Cuthbertson and E. Parr Metcalfe , B.Sc. ( Communicated by Professor Trouton , F.R.S. Received October 23 , \#151 ; Read November 15 , 1906 .
) ( Abstract .
) The present work forms a continuation of an enquiry begun by Mr. C. Cuthbertson* on the refractive indices of the vapours of elements not gaseous at ordinary temperatures .
The refractometer of Jamin was again employed , but tubes of fused silica were now used and other improvements introduced .
The following table exhibits the results obtained :\#151 ; Element .
Refractivities ( / x \#151 ; 1)106 .
A = 6562 .
X = 5893 .
X = 5630 .
X = 5540 .
A = 5460 .
A = 5183 .
Potassium negative __ _ negative Zinc 1960 2050\#151 ; 2065 2020 \#151 ; 2150 ?
2070 Cadmium 2675 2675 \#151 ; \#151 ; 2725 2780 Mercury 1799* 1866 \#151 ; \#151 ; 1882 1885 Arsenic \#151 ; 1552 \#151 ; \#151 ; 1579 Selenium 1520 1565 \#151 ; 1560 1570 Tellurium ... ... 2370 2495 2620 I * Subsequent work by Mr. Metcalfe renders it probable that the true value is 1J to 2 per cent , higher than this.\#151 ; January 24 , 1907 .
* 4 Phil. Trans./ A , vol. 204 , p. 323 , 1905 .

