]\gt ; Form of the Curve of for the emitted by Black Body .
By Prof. A. WILSON , F.R.S. , , London .
( Received January 4 , \mdash ; Read February 25 , 1909 .
) The view adopted in the paper is that the radiation from a black body is an ular disturbance subject to statistical laws .
This view is , I , that now fenerally held.* It is shown that these laws can be deduced from the observed distribution of energy in the spectrum and that they enable the general racter of the disturbance to be described .
At any point in an actual of the radiation we have a number of simple ] lonic vibrations of arbitrary phases and various amplitudes .
The radiation may be arded as the suln of all siulple ations in the spectrum .
The component vibrations may be arded ils continually , but the distl.ibution of energy the , }when averaged ovel a short interval remains Suppose that compound a lumber n of simple harmonic vibrations all along the same line and of the same amplitude , but phases distributed at random .
It is easy to calculate by the method by Lord the chance that the displacement at any instant selected at random will lie between } nits , say and It is ( where the amplitude of each of the COIll ) onents .
If compound a of such sets of rations and in th set denote the amplitude , the number by , and the by the expression for the chance of a iven displacement is Let the in a vibration be , and let .
Then in the spectrum of plane polarised radiation let denote the energy between and , so that corresponds with and Hence *Cf .
Larmor , 'Phil .
Mag vol. 10 , p. 574 , 1905 ; Rayleigh , 'Phil .
bIag vol. 11 , p. 123 , 1906 .
Theory of Sound , ' vol. 1 , p. 39 .
Prof H. A. Wilson .
The Statistical [ Jan. 4 , is the chance that the displacement at any instant in the radiation lies between and If denotes one of the component vibraticns , then the amplitude of .
Hence the chance that lies between and is whele .
The correexpressions for unpolarised radiation can easily be obtained in the usual way , but they have no special advantage .
It is more convenient for some purposes to consider as a function of distance along the path of the radiation , which will be denoted by .
Then , so that for is equal to Consider the average number of times per centimetre that the curve , representing the radiation , cuts the line constant .
Evidently is iven by the equation* for is the length per centimetre in pl occupied by the curve .
Hence .
Hers may be called the mean wave-length of the radiation .
Let denote the average number of maxima in per square ' centimetre at a distance from the axis , and denote the number of minima .
The number of times per ntimetre that vanishes is independent of , hence Also .
Hence and The number of zeros per centimetre in can be shown , in the same way as for , to be ; hence or is assuming that there is no correlation between and , which can easily be verified .
1909 .
] Forra ojthe urue , etc. Thus the distribution of maxima and minima can be determined from the .
In the same way the distribution of maxima and minima and Che number of zeros per centimetre for any of the derivatives of can ) found if desired .
In order to calculate the relative values of the 's it will be convenient to take the distribution of energy in the spectrum to be represented by Planck 's formula where denotes the absolute temperaCure and A and are constants whose approximate values are known .
This ives , putting The value of is unknown , so we can only calculate relative values of the .
The table gives the values ] and The larger values of are iyen almost exactly by the formula .
The mean wave-length is equal to , hence .
The wave-length , at which the energy in the spectrum is a maximum , is iven by the well-known equation .
Hence The total number of maxima and minima in is very ] double the number of zero values , and the number of points of inflexion is a little less than double the number of maxima and minima .
Substituting the values found for , and in the expression for we get The Statistical of the ) of Oscillation , etc. following table ives some values of The corresponding values of can be by cbanging ?
into The distribution of the zero values of or its derivatives along can easily be obtained .
If is the mean number of zeros per centimetre , then the number of spaces between two zeros reater than is , where is the whole number of spaces considered .
shows a curve drawn so as to have approximately the statistical properties determined above .
The spaces between the zeros maxima and minima and f , he amplitudes at the maxima and minima were selected at random from collections having the proper distributions .
It is important to consider under circumstances such a curve will resemble the actual curve showing the displacements in a stream of radiation .
Suppose the black body consists of a small hole in one side and an image of this hole is focussed on the slit of a spectrometer .
The disturbance at any point on the slit is the sum of the disturbances coming from an enormous number of .
electrons on the walls of the inside of the box .
Suppose each electron emits a series of more or less separate impulses ; then if the number of electrons concerned is big enough , the displacement at a point on the slit will be at any instant the sum of a large number of displacements each due to one electron .
In this case the displacenlent will practically never be zero , because no impulses have arrived , but only when the sum of the displacements in the impulses is zero .
This is the case to which the theory given above is intended to apply .
In this case , if a large number of short lengths of the displacement curve were selected at random , then most a great many of them would resemble the curve shown in fig. 1 .
In this sense this curve may be said to represent the character of the disturbance in the radiation .
If the distribution of the energy in the spectrum of the radiation emitted Photographic gement of Small Sotid Objects .
by a single electron could be examined it might be found , when ayeraged over a enough period , to be the same as that actually found , but we have no evidence as to this .
If the electron were freely moying about it would probably go all the possible states of the electrons , and so would probably the usual distribution of energy on the average over a enough period .
The theory given above is not intended to apply to this case .
The curve shown does equally well for any temperature proyided the scale of is taken to be proportional to the temperature and the scale of ?
/ inversely proportional to the temperature .
The theory given above applies of course equally well to the radiation emitted by any body or to radiation any kind of energy distribution in its spectrum .
Best Conditions for Photographic gement of Solid Objects .
A. MALLOCK , F.li .
S. ( Received January 26 , \mdash ; Read Febrnary 18 , 1909 .
) Having recently had to photograph some small solids on an enlarged scale , I was led to enquire into the conditions as to the angular aperture and focal length of the lens which would give the best definition in the picture obtained .
If the actual distance apart of the closest points in the object which must be separated in the picture is Jreater than several yths of the employed , the problem may be solved by geometrical optics .
Let principal focal length of the lens .
distance between some part of the object and the lens .
the focal length conjugate to diameter of lens .
resolving , e.g. , the distance between the closest points which it can separate .
nitude of the smallest detail the objects which to appear separated in the picture .

