Researches towards establishing a Theory of the Dispersion of Light .
By the Rev. BADEN POWELL , M.A. F.R.S. Savilian Professor of Geometry in the University of Oxford .
Received February 19 , -Read March 12 , 1835 .
Introductory Remarks .
THE phenomena of prismatic dispersion , as originally discovered by NEWTON , and since examined throughout a vast range of transparent bodies by succeeding philosophers , especially Sir DAVID BREWSTER , were principally considered with reference only to successive parts , or spaces , of the coloured spectrum , designated generally as the red , or violet , or mean rays .
The increasing precision of modern science has been evinced in the elaborate and justly celebrated researches of M. FRAUNHOFER , who , availing himself of the dark and bright lines to mark and designate distinct points of the spectrum , by prismlatic observations for ten different media , solid and liquid , has determined in each the refractive indices for seven principal rays , thus always absolutely identifiable .
We will use the term " definite rays " to signify the specific parts of the spectrum thus defined .
As to the law of the phenomena , the first notion of a simple proportionality was soon disproved .
The refrangibility was seen to vary considerably and irregularly for each ray and each medium ; and when FRAUNHOFER had assigned serieses of numbers as the accurate expressions of the varying refractive powers throughout the several spectra , the apparent absence of any law connecting these numbers was only rendered more palpable .
All that could be said was , that the numbers increased from the red to the blue end of the scale , and in a different way in each medium .
The first object of inquiry in the search after such a law , would be some other characteristic of the same definite rays , equally well determnined ; between which and the refractive index some connexion might possibly be found to subsist .
The only such characteristic , perhaps , is the length of the interval for each ray , the Newtonian fit , or the undulatory wave , which ( by whatever name it be called , ) has demonstrably a real existence in the nature of light ; and the value of which , for each of the definite rays , has also been determined by FRAUNHOFER , with his usual accuracy , from phenomena totally independent of refiaction , viz. , his very remarkable experiments , in which a spectrum absolutely pure and perfect is obtained without the intervention of any prism , from the interferences produced by a fine grating of parallel wires covering an object-glass .
The positions assumed by the successive rays , here depend on nothing but the lengths of their periods or waves simply as such ; and the intervals between them are precisely proportional to the differences of these lengths .
These lengths decrease from the red to the blue end of the spectrum .
We might search for some empirical law which should connect these two serieses of data , the one being some inverse function of the other ; but it would be more satisfactory should such a formula be supplied by any theory of light .
I shall not I trust be considered as assuming a controversial tone , if I observe that no researches directly suggesting any such formula have been published except those of M. CAUCHY , on the hypothesis of undulations .
In these , indeed , such a formula is not actually developed .
But in a paper in the London and Edinburgh Journal of Science* , in the former part of which I have offered a brief abstract of M. CAUCHY'S peculiar theory of undulations , some remarks upon it are given , including the deduction of a formula in which the relation between the length of a wave and the velocity of its propagation is precisely expressed ; this last quantity being in fact the same as the reciprocal of the refractive index .
Without entering any further into theoretical considerations , it will be admitted that such a formula , ( from whatever source derived , ) if found to supply anything like a representation of the law of nature , or a clue to guide us through the seeming disorder which prevails among the experimental results , would be entitled to attention .
It has therefore been my object , without reference to the support of a theory , to examine by means of this formula the relation between the index of refraction and the length of the period or wavefor each definite ray throughout the whole series of numerical results which we at present possess .
And it will become a matter of increasing interest to pursue observations on the indices of definite rays for a greater range of transparent media .
The present paper will be occupied with the discussion of the data already known ; and before proceeding to that discussion I will merely add , that whatever degree of interest may attach to the inquiry , the merit is due to Professor AIRY , in whose suggestion it originated .
General Observations on the Formula .
In the investigations in the paper above referred to , on substituting for the velocity of a wave expressed by -its equivalent - , the formula at once presents the relation between the index p and the length of a wave X. H , r , and n are quantities dependent on the nature of the medium ; r , by hypothesis , always being a sensible fraction of t : thus the formula becomes Isin ( r n ) Here the value of p will evidently vary with a change in the value of .
, or from one ray to another ; it will also vary with a change in the constants H , r , or n , that is , from one medium to another .
The mere inspection of the formula will suffice to show that it exhibits at least a general accordance with the obvious constitution of the prismatic spectrum in the greater dispersion of the blue end .
For , in general , as X is diminished , the arc --i increased , and consequently the ratio of the arc to its sine increases , or p increases .
And the variation in the value of this ratio , and consequently in that of ^ , for a given variation in x , is greater when the arc is greater , that is , when x is less .
Thus , towards the blue end of the spectrum , where x is least , the dispersion or expansion of the rays is greatest .
But we must proceed from these very general remarks to the more precise comparison of numerical values .
Comparison of Numerical Results .
In proceeding to apply the formula to actual calculation , we are met by several difficulties arising out of the peculiar form of the function .
The process is , in fact , reduced to finding arcs which shall fulfil the twofold condition of being themselves in the ratio of the values of h. , while they are to their sines in the ratio of the values of c. For this I have not been able to make any direct method available .
By indirect and tentative methods , however , and the assumption of arcs which were seen ( from a table of the lengths of arcs , ) to be nearly in the required ratio to their sines , I advanced by successive trials of greater or less arcs to more exact values .
Those for the two extreme rays were usually assumed in the first instance , and their ratios to their sines compared with the ratios of the refractive indices ; and these once brought to a sufficiently near accordance , a fundamental arc was obtained , from which those for the other rays were deduced on dividing by the corresponding value of ^ ; and the product of a constant coefficient multiplying the ratio of the arc and sine , which in theory ought to give the value of the refractive index , was compared with the index deduced from observation .
This will sufficiently explain the meaning of the several columns in the tabular statement of the results .
It must be borne in mind that the values finally adopted are still only approximative , and are open to further correction by repeating the process ; so that in all the cases here considered a still closer coincidence might probably be obtained were it thought desirable .
The fundamental data of these comparisons are ( as already said ) those very precise determinations of the value of x for the several definite rays named by the letters B , C , D , &amp;c. , obtained by FRAUNHOFER from the interference-spectrum ; and which 2K2 251 Sir J. HERSCHEL has justly characterized as data of the utmost value in the theory of light * .
These values are as follows : Ray .
Value of .
B -00002541 C -00002422 D -00002175 E -00001945 F -00001794 G -00001587 H -00001464 With these values I have gone through every case of a refractive index for a definite ray at present known , that is , for every one of these seven definite rays in each of the ten substances whose refractive energy for the different rays was examined by FRAUNHOFER .
The following tabular statement gives the comparison between the refractive index for each ray in each medium , as given by FRAUNHOFER'S observations in the first column , and as resulting from the formula of theory ( adopting his independent determinations of the values of ) in the last ; whilst in the intermediate columns the elements of the calculation are exhibited .
Flint glass , No. 30 .
FRAUNHOFER .
Assumed values .
Calculated vaCalculated vaRay .
Observed values frn Rtio ( aC )lue of of Z. t\sine const .
x\sine/ ' B 1-6236 16 0 1-0131 1-6239 C 1-6255 16 17 1-0135 1-6246 D 16306 18 15 1-0172 1-6305 E 1-6373 20 22 1-0214 1-6373 F 1-6435 22 8 1-0252 1-6434 G 1-6554 25 2 1-0325 1-6551 H 1-6660 27 9 1-0393 1'6660 const .
= 1-6033 Flint glass , No. 3 .
FRAUNHOFER .
B 16020 15 20 1-0120 1*6000 C 1-6038 16 5 1-0133 1-6039 D 1-6085 17 55 1-0164 i 16079 E 1-6145 19 59 1*0206 1-6145 F 1-6200 21 42 1-0243 1-6204 G 1-6308 24 33 1-0312 1-6313 H 1-6404 26 39 1'0369 1*6404 const .
= 1-582 Crown glass , M. FRAUNHOFER .
B 1-5548 12 19 1-0077 1-5548 C 1P5559 12 55 1.0085 1-5561 D 1-5591 14 23 1-0106 1-5593 E 1-5632 16 5 1-0133 1-5634 F 1-5667 17 26 1-0156 1-5671 G 1-5735 19 42 1P0199 1-5738 H 1*5795 2 22 1-0235 1-5792 const .
1-543 Crown glass , No. 13 .
FRAUNHOFER .
B 1-5243 1 118 1-0065 1-5243 C 1-5253 11 51 1*0071 1-5252 D 1-5280 13 12 1*0089 1-5279 E 1-5314 14 46 1-0112 1-5314 F 1-5343 16 0 1-0131 1-5343 G 1-5399 18 5 1-0168 1-5399 H 1-5447 19 37 1.0198 1-5444.const .
= 1-5145 Crown glass , No. 9 .
FRAUNHOFER .
B 1-5258 11 18 1-0065 1-5259 C 1-5269 11 51 1-0071 1-5269 D 1-5296 13 12 1-0089 1-5296 E 1'5330 14 46 1'0112 1-5332 F 1-5360 16 0 1-0131 1-5360 G 1-5416 18 5 1-0168 1-5416 H 1-5466 19 37 1.0198 1-5462 const .
= 1P5162 253 Conclusion .
'Upon comparing the numbers above given as resulting from theory and from observation , and bearing in mind that the assumptions of the constants on which the calculation depends are but tentative and approximative , and open to further correction , it will , I think , be allowed that the coincidences are quite sufficient to permit us to regard the formula as a very close representation of the law of nature .
We are thus , I think , justified in concluding , that for all the substances examined by FRAUNHOFER , viz. for four kinds of flint glass , three of crown glass , for water , solution of potash , and oil of turpentine , the refractive indices observed for each of the seven definite rays are related to the lengths of waves for the same rays , as nearly as possible according to the formula above deduced from M. CAUCHY 's theory .
Thus , then , for all the media as yet accurately examined , the theory of undulations ( as modified by that distinguished analyst , ) supplies at once both the law and the explanation of the phenomena of dispersion .
Oxford , February 17 , 1835 .
Oil of turpentine .
FRAUNHOFER .
Assumed values Calculated values Ray .
Observed values ofrn Ratio ( no ro of.o af of IA . ?
sine const .
X. Csi-ne ) B 1-4705 12 25 1-0078 1-4703 C 1-4715 13 1 1-0086 1-4715 D 1-4744 14 30 1-0107 1-4746 E 1-4783 16 13 1-0135 1-4786 F 1-4817 17 35 1-0159 1-4821 G 1-4882 19 51 1-0203 1-4886 H 1-4939 21 32 1-0239 1-4938 const .
=1-459 Solution of potash .
FRAUNHOFER .
B 1-3996 10 34 1-0056 1-3999 C 14005 11 5 1-0062 1-4008 D 1-4028 12 20 1-0077 1-4029 E 1-4056 13 10 1-0088 1-4044 F 1-4081 14 57 1-0114 1-4080 G 1-4126 16 55 1-0147 1-4126 H 1-4164 18 20 1-0173 1-4162 const .
=1-3922 Water .
FRAUNHOFER ( two experiments ) .
B 1-3309 9 54 1-0050 1-3309 C 1-3317 10 25 1-0055 1-3315 D 1-3336 11 36 1-0068 1-3333 E 1-3358 12 57 1-0085 1-3355 F 1-3378 14 3 1-0101 1-3376 G -1-3413 15 51 1-0129 1-3413 H 1-3442 17 11 1-0151 1-3443 const .
1-3243 I 254

