]\gt ; A Suggestion as to the Origin of Black Body diation .
By GEORGE W. WALKEB , A.B.C.Sc .
, M.A. , F.B.S. , formerly ellow of Trinity College , ( Received October 15 , \mdash ; Read November 13 , 1913 .
) The discussion on " " Radiation\ldquo ; in Section A at the recent British Association in was one of profound interest .
Some of the remarks made then suggested the investigation which I now describe .
Planck 's formula*for the emission function is .
This formla represents bhe observations for short and ] waves at various temperatures with considerable closeness ; and , if it is the correct expression , it is held by some to prove that the classical equations of dynamics and electrodynamics at fault .
As I do not think the classical equations are in much danger if properly applied , I have endeavoured to trace the dynamical explanation of the experimental data .
There must be many formulae which will express the data as well as Planck 's form .
In searching for such an expression of dynamical , I went back to the equations for the motion of a charged sphere which I established on dynamical principles .
A clue to a solution very soon appeared , and without further remark , at present , I will state the formula that was tried .
If the emission within a range is expressed by it will be remembered that , in order to satisfy Stefan 's and Wien 's laws , the expression must be a function of the product Accordingly , I selected the formula as a function which satisfies the following conditions:\mdash ; ( 1 ) It gives Stefan 's law that the total radiation ries as , since and .
( 2 ) It gives Wien 's law that the maximum radiation at any temperature occurs when constant .
In our formula 'Theorie .
Warmestrahlung , ' 1906 , p. 157 .
'Phil .
Trans 1910 , p. 152 .
Mr. G. W. Walker .
( 3 ) It gives the condition that the maximum radiation at any temperature for this wave-length varies as .
In our formula ( 4 ) It gives Lord formula for long waves These conditions are required by the experimental work of Kurlbaum , Wien , , Lummer and Pringsheim , and Rubens .
The remaining point is therefore whether the formula will fit with the lesnlts of Paschen or Lummer and Pringsheim for short wave-lengths .
I take Lummer and data as typical .
It soon appeared that one take , and then takes the simpler form We have at once from data where the units are in microns ( 1 cm and in rade degrees absolute .
In computing the ordinates for any specified value of , it is convenient to note that for any eIlgth equal to or , the function varies as .
The are the numerical values of the function .
I take the experimental of Lummer and at temperature 1450o absolute , for which is practically 2 , and choosing the scale to give a maximunl ordinate of 73 units as shown in their diagram , the curve from our formula was calculated , and is shown in fig. 1 .
It fits the whole experimental curve excellently .
We have thus obtained a formula which fits all the data as well as , if not better than , Planck 's .
It is empirical , and there may be many others .
But Verh . .
Deutsch .
Phys. Gesell 1899 , , .
217 .
A Suggestion as to the Origin of Black Body Radiation .
395 in form it ests what we may expect from a dynamical system with very heavy damping .
The equation where ?
and for the motion of an electron , has been much used by Lorentz , to whom we hs in owe it , and by Planck .
But it is only an nate equation , although for many purposes a very good one .
I have shown*that more correct expressions for the motion are given ) the two equations where is the value at of the lnction which defines the state of the field in the aetber , 'Phil .
Trans .
cit. Mr. .
W. Walker .
The equations were obtained first for a conductor , and were afterwards shown to be vely nearly true for an insulating sphere .
Let us suppose that the restoring force is linear , so that , and we then Thus the free motion is determined by where Hence is a root of the equation If , we get of which the roots are , squares of , or where is the frequency .
would agree with a calculation by Lorentz .
But experiments are against the supposition that the intrinsic mass of an electron , is zero , and Kaufmann 's numbers suggest that is of the same order as .
For a positiye particle it is generally agreed that is small .
Hence , retaining , we get two pairs of complex roots .
The approximate values squares of are where and These equations may be arded as applicable to the behaviour of a molecule made up of a heavy , and so comparatively stationary , positive particle with a electron revolving round it .
The vibrations are of two types .
In the first , which closely rees with A Suggestion as to the of Black Body .
397 Lorentz ' result , the damping is for optical purposes small , and the reduction of amplitude in the time between two collisions of a molecule with its neighbours would be nall .
But in the second type we have very different .
The frequency is enormous and the damping so for optical purposes that the reduction of amplitude to a small fraction of its original value may be regarded as instantaneous .
Just after each encounter the amplitudes of the two types may be regarded as comparable , but before the next encounter the second type will have been suppressed by almost stantaneous radiation , while the first type will not have suffered any great reduction .
Have we not here a clue to the " " quantum\ldquo ; theory and the characteristic Bontgen radiation ?
In a detailed discussion it will be necessary to take account of the radiation the positive particle also .
Let us now consider the steady radiation from an electron in which the motion is governed by the equations and is maintained by a purely mechanical force .
The mean rate of radiation , viz. , , I find to be where Tbis result is equftlly applicable to a positive particle with appropriate values of , and .
A similar result hold when the joint action of a positive and Cive combination is considered , and ?
are a combination of the intrinsic and electric inertia ter1ns , hile would now be a linear quantity detining the radius of the orbit .
The coefficient of is , as far as or enters , simply eneralised form of and I have little doubt that it could be fitted with the data .
There are always three positive values of for which is a maximum .
These are in the vicinity of , and The corresponding maximum intensities are in the ratio nearly .
If we choose so that the first comes in the observed range of wave398 A Suggestion to the Origin of Black Body Radiation .
thso , the other two are in the ultra-violet , and if is small , they are of slight importance as compared with the first .
If , and only in so far as , Wien 's and Stefan 's laws are true , we require to suppose that varies as and varies as , and the expression is then a function of .
The reconciliation of these with each other and with the virial theorem of Clausius is a matter of difficulty , but as the experimental evidence is that either lnw is strictly true , the matter may well rest for the present .
These considerations , and the present estimates of atomic magnitudes , lead me to suspect that black body radiation is determined not by the electron , nor by the positive particle alone , but by the joint action of the two .
My conclusions are : \mdash ; ( 1 ) That the experimental data can be well represented by a formula of dynamical type , of which I yiven one .
( 2 ) That Newtonian dynamic and the electrodynamics of are capable of giving an explanation of that formula .
[ Note added October 23 .
1913.\mdash ; Plallck 's radiation function has been applied by Einstein*and by to the explanation of the variation of atomic heats with tempel.ature .
By similar reasoning I find that if the radiation function varies as the typical term in the expression for the atomic heat varies as where is some definite temperature .
The following table ives the numerical values The result is similar to that required to explain the observations of Nernst .
] I would express my rateful thanks to Sir Joseph Larmor , who haS discussed this paper with me , and by his gestions has added so mnch to the manner of presenting the lesults .
'Ann . .
Phys 1907 , vol. 22 , p. 184 .
Sitzber . .
Berl .
Akad 1911 , p. 494 .

