2

Max Planck and the Law of

Blackbody Radiation*
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

1. The Early Work of Robert Gustav Kirchhoff

by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Towards the end of 1859 Gustav Robert Kirchhoff, Professor of Physics at the
University of Heidelberg, submitted two papers to the Prussian Academy concerning
questions of radiation. 1 One of them dealt with the explanation of the so-called
Fraunhofer lines, i.e., the dark lines observed in the solar spectrum (Kirchhoff,
1859a); it contributed, together with the experimental investigations carried out by
Kirchhoff and his colleague Robert Bunsen, to establishing the methods of chemical
spectral analysis (Kirchhoff and Bunsen, 1860). The other did not lead immediately
to any practical application; in it a rather specific property of bodies emitting
light and invisible heat radiation was expounded, namely the fact that the ratio
of emissivity to absorptivity must be the same for all bodies, provided a given
wavelength of the radiation is observed and the bodies have the same temperature
(Kirchhoff, 1859b). Kirchhoff concluded further:
The ratio of the power of emission to the power of absorption, e/a, common to all bodies,
is a function depending on the wavelength [of the radiation emitted or absorbed] and the

Invited Planck Lecture, delivered on 15 December 1975, at the Akademie der Wissenschaften
der DDR, East Berlin, to mark the 75th anniversary of Max Planck's discovery of the quantum
of action. Revised and enlarged version published in The Historical Development of Quantum
Theory (with Helmut Rechenberg, Springer-Verlag New York, 1982, pp. 24-59).
1
Gustav Robert Kirchhoff was born on 12 March 1824 in Konigsberg. From 1842 he studied
at the University of Konigsberg, especially under Franz Neumann. After receiving his doctorate
in 1847 he went to Berlin and became Privatdozent at the University of Berlin a year later. In
1850 he was called to Breslau as an Extraordinarius; there he met the chemist Robert Wilhelm
Bunsen (1811-1899). Bunsen, who moved to Heidelberg the following year, proposed (in 1854)
his friend Kirchhoff for the Professorship of Physics at the University of Heidelberg, which had
become vacant when Philipp von Jolly left for Munich. Finally, in 1875 Kirchhoff was appointed
to the chair of theoretical physics at the University of Berlin, where he died on 17 October 1887.
Kirchhoff made numerous contributions to experimental and theoretical physics. He worked on
electricity, especially on the connection between electrostatic and electrodynamic concepts, and
derived a theorem which gives t h e distribution of the currents in a network (Kirchhoff's rules).
He worked on various other problems, such as the thermal conductivity of iron, reflection and
refraction from crystals, and the thermodynamics of solutions. However, his most important
researches concerned the emission and absorption of light; these were important not only for
physics but also for chemistry and astrophysics.

19
 20 The Golden Age of Theoretical Physics

temperature. At low temperatures this function assumes the value zero in the case of the
wavelengths of visible radiation, and values different from zero for larger wavelengths; at
higher temperatures the function [e/a] takes on a finite value also for the wavelengths of
visible rays. At the temperature for which this function (considered for the wavelength
of a given visible ray) ceases to be zero, all bodies begin to emit light having the colour
of this ray, except those [bodies] which possess a negligibly small power of absorption for
[rays of] this colour at this temperature; the larger the power of absorption, the more light
does a body emit. (Kirchhoff, 1859b, p. 726) 2

In a subsequent paper in Annalen der Physik Kirchhoff discussed t h e relation be-

tween emissivity and absorptivity in great detail; in particular, he introduced there
t h e concept of a completely black body, defining it as a b o d y which absorbs all
radiation falling u p o n it (Kirchhoff, 1860).
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

Although Kirchhoff's discussion of the emission a n d absorption of visible a n d

(invisible) heat radiation a p p e a r e d t o deal w i t h a n a b s t r a c t problem of only theo-
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

retical interest, m a n y physicists concerned themselves during t h e following decades

with t h e properties of the function $(A, T ) ,

^,T) = Qx, (!)

which expressed t h e ratio e/a, emissivity to absorptivity, for a given wavelength
of radiation, A, on the t e m p e r a t u r e T. ( T h e absorptive power of a body, o^, is
defined as t h e ratio of the intensity of the absorbed to t h e incident radiation for a
given wavelength A.) T h e reason for t h a t interest m a y b e easily u n d e r s t o o d if one
takes into account the general situation t o w a r d s the end of t h e nineteenth century.
At t h a t time, for instance, t h e illumination of t h e cities at night, b o t h by gas a n d
electricity, was introduced; evidently, for practical a n d economic reasons it became
necessary t o find t h e most efficient means of producing visible light. Second, since
t h e exploration of the e a r t h and the continents seemed to move towards t h e s t a t e
of completion, m a n turned his attention increasingly t o t h e study of celestial ob-
jects a n d stars. T h e investigation of the s p e c t r a of stars, in particular, seemed to
provide i m p o r t a n t hints concerning their properties, such as their surface temper-
atures. T h e latter could b e derived, in principle, by observing the distribution of
intensity of t h e radiation among t h e wavelengths a n d comparing t h e result with
2
Kirchhoff was not the first to discuss the relation between emission and absorption of (heated)
bodies. After the experimental work of John Leslie (1766-1832), establishing the proportionality
between the total emission and absorption of two bodies (Leslie, 1804), especially the Scottish
physicist and meteorologist Balfour Stewart (born on 1 November 1828 at Edinburgh, died on
19 December 1887 at Manchester) concerned himself with the problem. In a paper, entitled 'An
Account of Some Experiments on Radiant Heat' and read in March 1858 before the Royal Society
of Edinburgh, Stewart presented his empirical results claiming that the absorption of a heated
plate equals its emission for all kinds of radiation (Stewart, 1858). In this paper Stewart also gave
a theoretical proof for his finding; he argued, from the assumption that particles in the interior of
bodies emit equal amounts of radiation independently when at uniform temperature, that emission
and absorption of any type of heat radiation must be equal for any part of a body. His proof did not,
however, allow him to compare different materials and to postulate, as Kirchhoff did, a universal
function of temperature and wavelengths. A historical discussion of the respective achievements
of Stewart and Kirchhoff in radiation theory has been given by Daniel M. Seigel (1976).
 Max Planck and the Law of Blackbody Radiation 21

Kirchhoff's function $(A,T). Third, after the recognition of light as electromag-

netic waves one turned to explore systematically the full spectrum of these waves,
discovering quite new forms of radiation. Thus, Samuel Pierpont Langley extended
the range of observable heat radiation to include wavelengths of one-millionth of a
metre by his bolometer method (Langley, 1886) and Heinrich Hertz demonstrated
the existence of electromagnetic waves emitted by oscillating currents in open cir-
cuits, having wavelengths of the order of a metre (Hertz, 1888a). On the other
hand, the electromagnetic spectrum was also extended on the ultraviolet side to
shorter and shorter wavelengths, especially through the discovery of X-rays and of
7-radiation from radioactive substances, which took place shortly before the end of
the nineteenth century and in the beginning of the twentieth century, respectively.
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

2. Towards t h e Derivation of Wien's Law

by KAINAN UNIVERSITY on 01/10/15. For personal use only.

The discoveries mentioned above evidently required an immense improvement of the

experimental methods; completely new techniques had to be developed to register,
say, the long electromagnetic waves, and the precision and reliability of known tech-
niques had to be increased. In this development the physicists received help from
outside; because of the needs of growing industrial production in Europe and the
United States a strong pressure was exerted on governments and educational insti-
tutions to provide the means for such improvements of techniques and methods. On
one hand, laboratories were installed at universities and technical institutions were
established, equipped with modern instruments and apparatus; thus, for example,
the establishment of the Cavendish Chair and Laboratory at Cambridge University
was strongly motivated by the idea that the work performed in the Laboratory might
result in obtaining better and more accurate standards for industrial purposes. On
the other hand, for the same reason, special national institutes and laboratories were
also created, which had not only to control the standard weights and measures but
also to develop methods and standards for new technologies. The first of these insti-
tutes (or bureau of standards), the Physikalisch-Technische Reichsanstalt in Berlin,
came into being in 1887, with substantial help from the industrialist and inventor
Werner von Siemens, the founder of the electrical firm of Siemens una1 Halske.4 The
various precision methods developed at the Physikalisch-Technische Reichsanstalt
were combined towards the end of the nineteenth century to provide fundamental
data concerning the energy distribution in blackbody radiation — data on which a
theoretical understanding of Kirchhoff's function could be achieved.
Ever since Friedrich Wilhelm (William) Herschel had demonstrated the heat ef-
fects of infrared radiation in the very beginning of the nineteenth century, physicists
had been interested in the study of the radiation emitted by hot bodies. In order
to investigate the visible part of the spectrum, optical and, later on, photographic
3
T h e electromagnetic nature of X- and 7-rays was confirmed only much later, i.e., in connec-
tion with the discovery of their interference effects in crystal lattices (Friedrich, Knipping and
Laue, 1912).
4
For a history of the Physikalisch-Technische Reichsanstalt see the article of Karl Scheel (1913).
 22 The Golden Age of Theoretical Physics

methods were used. The latter were also found useful for the ultraviolet part. The
infrared region, the heat radiation in the original sense of the word, could be ob-
served by its temperature-increasing effect, which was objectively registered, say, by
a thermocouple. The experimental observation of infrared radiation was substan-
tially improved when Samuel Pierpont Langley invented the so-called 'bolometer,'
an instrument using the temperature-dependent change of the resistivity of platinum
wire blackened by carbon (Langley, 1881).5 With it Langley observed the radiation
emitted from heated copper, including wavelengths up to 5.3 fi; he demonstrated,
in particular, a definite displacement of the maximum of intensity with increasing
temperature of the copper — the temperature of the probe was varied between
330°C and 815°C — towards smaller wavelengths (Langley, 1886). Langley's mea-
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

surements were extended by Priedrich Paschen, then at the Technical University of

Hanover.6 Paschen entered the field in 1892 and concerned himself for several years
with various techniques of observing and measuring the intensity of heat radiation,
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

employing many different substances such as incandescent gases and solid matter.
Finally he obtained detailed results from several solid emitters including carbon,
copper oxide and platinum; these results converged to yield an empirical formula
for Kirchhoff's function, that is,

$(A,T) = C l A-Qexp(-^), (2)

where ciand c^ were constants and the negative exponent — a assumed the value
of about —5.5 (Paschen, 1896). Shortly before Paschen's publication of his empir-
ical law, Wilhelm Wien had arrived at the same result on the basis of theoretical
arguments (Wien, 1896).

S.P. Langley was born on 22 August 1834 in Roxbury, Boston, Massachusetts. He completed
his formal education on graduating from Boston High School in 1851. After travelling in Europe
(1864-1865), he became an assistant at the Harvard Observatory in 1865. In 1866 he was ap-
pointed assistant professor of mathematics at the U.S. Naval Academy and placed in charge of
the observatory. In 1867 Langley was made Director of the Allegheny Observatory and Professor
of Physics and Astronomy at Western University, Pennsylvania. From 1887 onwards Langley was
Secretary of the Smithsonian Institution in Washington, D.C. He died in Aiken, South Carolina,
on 27 February 1906.
Besides astronomical researches, especially on stellar spectra, Langley made aerodynamical
studies and designed, for instance, engine-driven airplanes. He received many honours and prizes,
including memberships in the U.S. National Academy of Sciences, and the Royal Societies of
London and Edinburgh. He was awarded the Rumford and Henry Draper Medals.
Louis Carl Heinrich Priedrich Paschen was born in Schwerin, Mecklenburg, on 22 January 1865.
He studied mathematics, natural science and physics at the Universities of Strasbourg and Berlin
from 1884 to 1888, receiving his doctorate from Strasbourg in 1888. In 1888 he became an
assistant to Johann Wilhelm Hittorf (1824-1914) at the University of Miinster; in 1891 he became
an assistant of Heinrich Kayser (1853-1940) in Hanover, where he received his Habilitation two
years later and held the position of a Dozent for 'Physics and Photography' from 1895. In 1901
Paschen was appointed Ordinarius (full professor) of physics at the University of Tubingen. He
stayed there, apart from a short period at the University of Bonn (1919-1920), until 1924, then
he was called to Berlin to succeed Emil Warburg as President of the Physikalisch-Technische
Reichsanstalt, where he remained from 1924 to 1933. He died in Potsdam on 25 February 1947.
Paschen worked primarily in spectroscopy, both on heat radiation and optical spectroscopy.
Under his leadership Tubingen became the 'mecca of spectroscopy.'
 Max Planck and the Law of Blackbody Radiation 23

Equation (2) describes the energy density distribution, p\, among the various
wavelengths of the radiation emitted by an idealized body, which Gustav Kirchhoff
had called a 'completely black' or just a 'black' body, i.e., a body which possesses the
property of absorbing all the radiation falling upon it (Kirchhoff, 1860, Section 1).
The first attempts to derive theoretically the energy distribution of blackbody ra-
diation, which is identical with Kirchhoff's function, i.e., <fr(A,T) = p\, were made
by Eugen Lommel (1837-1899), then Professor of Physics at the University of Mu-
nich, by Vladimir Alexandrovich Michelson (1860-1927), then at the University of
Berlin, and by Ludwig Boltzmann, then professor of physics at the University of
Graz. While Lommel had based his approach on a mechanical model describing
the vibrations in a solid body (Lommel, 1878), Michelson had used arguments from
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

kinetic gas theory, including Maxwell's velocity distribution of molecules, to arrive

at a formula for $(A, T) (Michelson, 1888).7 On the other hand, Ludwig Boltzmann
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

had based his treatment right away on Maxwell's electromagnetic theory of light
when he derived the law obtained empirically by Joseph Stefan, which accounted
for the total intensity of the radiation (both visible and invisible) emitted by a
heated blackbody (Stefan, 1879). Boltzmann had also used a connection between
radiation and the second law of thermodynamics that had been put forward sev-
eral years earlier by the Italian physicist Adolfo Bartoli. 8 Bartoli had considered, in
particular, the possibility of increasing (by an adiabatic process) the temperature
of a body situated in an enclosure through which heat radiation passed; this adia-
batic process consisted in the reduction of the volume of the enclosure, and he had
found that, in order not to violate the second law of thermodynamics, one had to
take into account the pressure exerted by the heat radiation on the body (Bartoli,
1876). In his first note analyzing Bartoli's result, Boltzmann had arrived at the
conclusion that the existence of a radiation pressure of magnitude p = hp, where p
is the radiation density, was consistent with Stefan's law (Boltzmann, 1884a). In a
second note Boltzmann had then sharpened his conclusion to state the following: if
the second law was valid and if heat radiation — like any radiation — possessed a
pressure of the above magnitude, Stefan's law, i.e.

P = AT\ (3)

with A a constant, would follow immediately (Boltzmann, 1884b). 9

7
For a detailed discussion of Michelson's result, see Kangro, 1970, Section 2. Let us just mention
here that from his function followed the constancy of the product, A ^ a x • T, where A m a x is the
wavelength having the maximum intensity and T the absolute temperature.
8
Adolfo Bartoli, born on 19 March 1851 in Florence, studied physics in Pisa and Bologna, and
became Professor of Physics at the University of Sassari (1878) and the Technical Institute of
Florence (1879-1886), then Professor and Director of the Observatory at the University of Catania,
and finally Professor at the University of Pisa. He died on 18 July 1896 in Pavia. Bartoli worked
on various problems, including the specific heat and the dissociation of water and on electrolysis.
He demonstrated the existence of radiation pressure.
9
T o find this result, Boltzmann used the equation, Tdp — pdT = pdT, which can be derived from
the second law of thermodynamics. (It follows from the fact that dQ/T is a total differential,
 24 The Golden Age of Theoretical Physics

Bartoli and Boltzmann's idea of bringing both electrodynamics and thermody-

namics into the treatment of heat radiation was picked up and followed further
by Wilhelm (Willy) Wien, who had served as an assistant at the Physikalisch-
Technische Reichsanstalt since 1890.10 In his first publication on this subject, enti-
tled ' Uber eine neue Beziehung der Strahlung schwarzer Korper zum zweiten Haupt-
satz1 (Wien, 1893a), he immediately extended the earlier treatments: instead of
talking about the energy of the full spectrum — as Boltzmann had done — he
was interested in the energy distributed among given wavelengths. 11 Wien stud-
ied, especially, the enlargement of the energy density of radiation caused by the

where dQ is the sum of the changes in the internal energy and the external work.) In this equation
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

he substituted the relation p= ^p and integrated the resulting equation, putting the constant of
integration equal t o zero, from which Eq. (3) followed.
Wilhelm Carl Werner Otto Fritz Franz Wien was born on 13 January 1864 in Gaffken near
Fischhausen, East Prussia. After attending the Gymnasium in Rastenburg and Konigsberg, he
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

studied mathematics and physics in Gottingen (1882), Berlin (1882 to the winter semester 1883-
1884), Heidelberg (summer 1884) and again Berlin (winter semester 1884-1885 to winter semester
1885-1886), receiving his doctorate in 1886 under Helmholtz with a thesis on a problem of the
diffraction of light by sharp edges. Wien then worked for several years on the agricultural es-
tate (Landgut) of his father until he received the appointment as Helmholtz' assistant at the
Physikalisch-Technische Reichsanstalt in 1890. He obtained his Habilitation two years later with
a theoretical work on the 'localization of energy.' (Wien considered the concept of 'localization of
energy' in connection with Poynting's theory of the propagation of energy in the electromagnetic
field and the subsequent work of Heinrich Hertz.) Wien discussed the flux of energy in hydro-
dynamics, in elastic bodies and finally electrodynamics, including radiation theory. (See Wien,
' Uber den Begriff der Lokalisierung der Energie' ('On the Conception of Localization of Energy'),
1892.) In 1896 he was called to the Technische Hochschule, Aachen, as Extraordinarius for physics
(succeeding Philipp Lenard); in 1899 he succeeded Otto Wiener as Ordinarius in Giessen, and a
year later he moved as Ordinarius to Wurzburg (this time succeeding Wilhelm Conard Rontgen).
Finally in 1920 he obtained the Chair of Experimental Physics at t h e University of Munich, again
succeeding Rontgen. He died in Munich on 30 August 1928.
Wilhelm Wien was one of those rare twentieth-century physicists who worked as a specialist
both in experimental and theoretical physics. His researches on blackbody radiation won him the
Nobel Prize in Physics in 1911. Wien worked in thermodynamics and hydrodynamics and made
pioneering experimental studies of the electric and magnetic deflection of canal and cathode rays
(the latter contributed to the discovery of the electron); he also worked on X-rays and on the
recombination of ions.
11
Wien became interested in the problem of heat radiation on his own. T h u s he spoke in his
autobiography (written in 1927, and published posthumously in 1930) about the freedom which
he had at the Reichsanstalt to deal, besides his official task of establishing a standard of light
intensity, also with theoretical topics of his choice, such as water waves and cyclones. He wrote:
The independence, which I soon achieved in scientific work, immediately bore fruits. I
turned to the field of heat radiation and succeeded in discovering in it — without really
great effort — new laws, which received the approval of the scientific world. Even today
it gives me great satisfaction that my first paper on the displacement law [Wien, 1893a]
was presented by Helmholtz — though after some resistance — to the Berlin Academy.
He [Helmholtz] told me that he had thought earlier that the radiation could not be
treated thermodynamically, but that he had been convinced [by my work] that I was
right. I did not understand Helmholtz' reservations, for in my opinion radiant heat was
an integral part of heat itself and had to satisfy the same laws. Only later did I come to
know that Lord Kelvin had spoken much more definitely against my theories, for he said:
"Thermodynamics are going mad." I gave the general formulation of my ideas on the
properties of the heat radiation in my paper on ' Temperatur und Entropie der Strahlung'
('Temperature and Entropy of Radiation' [Wien, 1894]). Apart from a few minor points,
it has been fully accepted into theoretical physics. (Wien, 1930, pp. 16-17)
 Max Planck and the Law of Blackbody Radiation 25

following two processes: the increase of the temperature and the adiabatic decrease
of the volume of the enclosure containing the radiation; and he demanded that
both processes should lead to the same energy distribution among the wavelengths
if the same final temperature were reached. He achieved this goal after taking into
account Doppler's principle, that is, the fact that A, the wavelength of the emitted
radiation, depends on the velocity of the source. In particular, he obtained the
result that the densities, p\Q and p\, associated with Ao and A, the wavelengths
before and after the volume change, respectively, were related as p\j' p\0 = (AoA)4;
hence, due to the Stefan-Boltzmann law, Eq. (3), there followed the equation

TX = TQXQ = const. (4)

The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

Equation (4) expresses the fact, which later on came to be called 'Wien's displace-
ment law,' that for blackbody radiation the product of temperature and the related
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

(through the adiabatic process) wavelength remains a constant; a special example

of this law is provided by taking for A the wavelength, A max , having the maximum
intensity or energy density at the given temperature. 12
Wien continued to investigate the properties of blackbody radiation during the
following years. Especially, he used the concept of entropy to rederive Eq. (4) and
to obtain an inequality, that is,

const. ,„
^ - * 5 - > (5)

for p\, the energy density of blackbody radiation in the interval of wavelength be-
tween A and X + dX (Wien, 1894). A couple of years later he arrived at an explicit
law for px by applying the 'molecular hypothesis' in addition to thermodynamic
arguments (Wien, 1896). In particular, Wien started by assuming that blackbody
radiation was emitted by molecules obeying Maxwell's velocity distribution; that
is, the number of molecules having the velocity v was proportional to the quan-
tity i;2 exp(const. v2/T), where T denotes the absolute temperature. Further, he
assumed that A, the wavelength of radiation emitted by a molecule, was a function
only of its velocity v and vice versa. In this way he obtained the result

px = F(X)exp(-^-) , (6)

where F(X) and /(A) are functions of the wavelength A. By applying the displace-
ment law, Eq. (4), Wien determined the function in the exponent to be

12
I n the latter sense, Wien's law was used in future. The name 'displacement law' (' Verschiebungs-
gesetz') first occurred in a paper by Otto Lummer and Ernst Pringsheim six years later (Lummer
and Pringsheim, 1899b, p. 219).
 26 The Golden Age of Theoretical Physics

and from Stefan-Boltzmann's law, Eq. (3), he derived that

FW = %- (6b)

Hence Wien's Eq. (6) agreed with Friedrich Paschen's empirical law, Eq. (2), pro-
vided one could identify p\ with Kirchhoff's function $(\,T), Eq. (1), and take
the value 5 for Paschen's power exponent; thus it reproduced the observed data
perfectly.
In June 1896 Willy Wien left Berlin to take a professorship at the Technische
Hochschule of Aachen, at a time when the Physikalisch- Technische Reichsanstalt
became increasingly involved in the preparations for absolute measurements of the
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

blackbody radiation law. It was very fortunate that Max Planck, who had succeeded
Gustav Kirchhoff as Professor of Theoretical Physics at the University of Berlin,
at that time became a Haustheoretiker (resident theoretician) of the experimental-
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

ists working on blackbody radiation. Max Karl Ernst Ludwig Planck was born in
Kiel on 23 April 1858, the son of a law professor at the university of that city; he
had received his early education in Kiel and Munich and had studied physics and
mathematics at the Universities of Munich (1874-1877) and Berlin (1877-1878). 13
Among his professors were Philipp von Jolly (1809-1894) in Munich and Hermann
von Helmholtz (1821-1894) and Gustav Kirchhoff (1824-1887) in Berlin. Planck
obtained his doctorate, summa cum laude, from the University of Munich in 1879
with a thesis entitled ' Uber den zweiten Hauptsatz der Warmelehrd (Planck, 1879),
and became Privatdozent in Munich the following year. In 1885 he accepted a call
to an Extraordinariat (associate professorship) at the University of Kiel, and four
years later he moved to the University of Berlin, where he was promoted to a full
professorship of theoretical physics in 1892. Planck devoted his early scientific ca-
reer to the investigation of one general topic, the second law of thermodynamics,
especially the concept of entropy and its application to problems of physical and
chemical equilibrium, such as phase transitions and electrolytic dissociation. 14 His
first published papers already exhibited the characteristic features of his later work:
on one hand, he carefully worked out the details of his theories and calculated
results that could be compared immediately with the available experimental data
(see, e.g., Planck, 1881, 1890); on the other, he put great emphasis on clear defini-
tions of the fundamental concepts. Having been deeply influenced by the writings
of Rudolf Clausius, he had sought in particular to establish the 'principle of the

13
Planck's father Johann Julius Wilhelm von Planck, came from an academic family (both his
grandfather, Gottlieb Jakob Planck, and his father, Heinrich Ludwig Planck, were professors of
theology at the University of Gottingen). His second wife, E m m a Patzig, came from Greifswald.
The Planck's had seven children, Hugo and E m m a (from Johann Julius Planck's first wife), Her-
mann, Hildegard, Adalbert, Max and Otto (from his second wife); thus Max Planck was the fourth
son of Johann Planck.
14
Planck did fundamental work on these fields, i.e., both on phase transitions and electrolytic
dissociation. However, his contributions were not the first: on phase transitions, Josiah Willard
Gibbs had published his pioneering papers several years earlier, while on dissociation the Swedish
chemist Svante Arrhenius was slightly ahead of Planck.
 Max Planck and the Law of Blackbody Radiation 27

increase of entropy' in thermodynamics on the same level as the law of conserva-

tion of energy. In this endeavour he had pointed out the crucial role of irreversible
processes, i.e., the fact that in nature process occur, such as the conduction of heat,
which cannot be reversed completely. Planck's analysis of mechanical and ther-
modynamic concepts had brought him into disagreement with Wilhelm Ostwald
and the partisans of 'energetics' ('Energetik'). For example, at the Liibeck Assem-
bly of Natural Scientists (Naturforscherversammlung) in 1895, where Ostwald and
Georg Helm had defended the views of the energeticists and Boltzmann had at-
tacked them, Planck had been on Boltzmann's side. 15 However, unlike Boltzmann,
who had argued against Energetik on the basis of the molecular hypothesis and the
kinetic theory of matter, Planck had used thermodynamic reasoning in order to
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

criticize Ostwald's concept of 'volume energy.' Indeed, Planck had adopted a very
cautious attitude towards the molecular hypothesis and, in a paper entitled lGegen
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

die neuere Energetik' ('Against the New Energetics'), had declared concerning this
point: 'I do not intend, at this point, to enter the arena [on behalf of] the mech-
anistic view of nature; for that purpose, one has to carry out far-reaching and, to
some extent, very difficult investigations' (Planck, 1895b, p. 73).
At the time of the controversy about Energetik Planck had turned his attention
to what was a new field of investigation for him: heat radiation. Several reasons
may be cited why he had become interested in this field. First, there was the gen-
eral interest of many physicists in the phenomena of electromagnetic waves after
Heinrich Hertz' successful experiments. 16 A second reason was Planck's concern
with the importance of thermodynamic arguments in electromagnetism. Thus, for
example, in his inaugural address (Antrittsrede) to the Prussian Academy on 28
June 1894, he had expressed the hope 'that we can obtain a closer understanding
also of those electrodynamic processes, which are directly caused by the [action
of] temperature and which show up especially in heat radiation, without having
to follow the laborious detour through the mechanical interpretation of electricity'
15
Later on Planck recalled the discussions about Energetik at t h e Liibeck Naturforscherversamm-
lung in the following words:
It is evident t h a t this fight, in which especially Boltzmann and Ostwald opposed each
other, was carried out rather spiritedly; it also led to some drastic effects since both
opponents were well-matched in quickness and wit. In it, according to what I said
earlier [concerning my position], I could only play the role of a second to Boltzmann,
whose services were not only not recognized but even not liked by him [Boltzmann].
(Planck, 1948a, p. 20)
16
T h e general interest in Maxwell's electrodynamics and its consequences may be discerned, for
example, by the great number of lecture courses on it at the University of Berlin during the early
1890s. Among others, courses were given by Willy Wien (in the summer semester 1892 and winter
semester 1892-1893), Max Bernhard Weinstein (1852-1918) who had translated Maxwell's Treatise
on Electricity and Magnetism into German (in the winter semester 1891—1892), and Heinrich
Rubens (on the experimental foundations of the subject in the winter semester 1894-1895). Max
Planck also lectured in Berlin regularly on the theory of electricity and magnetism (e.g., winter
semesters 1889-1890, 1892-1893, and 1895-1896) (see Kangro, 1970, Sections 6.1.2 and 6.1.3).
In this connection, it is worth noting that Planck gave the memorial address on Heinrich Hertz
on 16 February 1894 at the Berlin Physical Society, in which he emphasized Hertz' role in the
development of Maxwell's theory and electromagnetic waves (Planck, 1894a).
 28 The Golden Age of Theoretical Physics

(Planck, 1894b, p . 643). Evidently, Planck h a d t h e occasion to p a r t i c i p a t e in m a n y

discussions between his Berlin colleagues, such as Willy Wien a n d Heinrich Rubens,
who worked actively on t h e problems of heat radiation. And, before starting his
theoretical investigations, he knew t h a t detailed experiments to explore t h e proper-
ties of blackbody radiation were being m a d e ready at the Physikalisch- Technische
Reichsanstalt. T h u s he reported later in his scientific autobiography:
By the measurements of [Otto] Lummer and [Ernst] Pringsheim at the Physikalisch-
Technische Reichsanstalt, performed to investigate the spectrum of heat radiation, my
attention was directed to the theorem of Kirchhoff, stating that in an evacuated cavity —
bound by totally reflecting walls and containing emitting and absorbing bodies that are
completely arbitrary [in shape, material and temperature] — a state will be established in
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

course of time, in which all bodies assume the same temperature and in which all proper-
ties of the radiation [contained in the cavity] — even its spectral energy distribution — do
not depend on the structure and composition of the bodies, but solely on the temperature.
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

This so-called normal energy distribution [of the radiation thus obtained in the cavity],
therefore, represents an absolute quantity; and, since the search for the absolute always
appeared to me to be the most beautiful ("schonste") task of research, I eagerly started
to deal with it. (Planck, 1948a, pp. 23-24)

M a x Planck was aware, of course, of Wien's law, Eq. (6), which represented the ex-
isting observations extremely well. However, t h e derivation given by Wien, t h o u g h
it employed t h e r m o d y n a m i c arguments, did not satisfy Planck's t a s t e completely.
He hoped to arrive at t h e same or a similar equation in a more systematic way,
using fewer hypotheses t h a n his predecessor.
Planck presented the first paper, in which he concerned himself w i t h t h e prop-
erties of electromagnetic radiation, t o the Prussian Academy in its session of 21
March 1895. I n this work he studied the processes of absorption a n d emission of
radiation or, as he called it, a 'resonator' (Planck, 1895a, p. 296). 1 7 He continued

1
Planck referred in his paper first to the 'secondary conductor, whose characteristic period al-
most coincides with the period of the primary wave' and, therefore, will be 'excited by resonance
to perform electric oscillations, the more so the less the periods differ from each other' (Planck,
1895a, p. 289). With this statement he had in mind Heinrich Hertz' experimental setup consisting
of 'primary and secondary conducting systems' ('primdre und secunddre Letter,' Hertz, 1888a,
p. 552), which Hertz had used to determine the propagation of electromagnetic waves; that is, the
waves were created in the primary system and directed in the secondary system. In a later paper
Hertz had also given the theory of his systems; he represented both the primary and the secondary
by a conductor of length I, in which an electric charge of magnitude e performed harmonic os-
cillations, and he investigated the properties of the electromagnetic waves that were emitted by
the primary system and absorbed and reemitted by the secondary system (Hertz, 1888b). Planck
quoted these results of Hertz when establishing his theory of absorption and emission of electric
waves by resonance (see Planck, 1895a, p. 290, footnote 1).
It should be noted that Planck later on used the words 'resonator' and 'oscillator' synonymously
for the elementary systems which absorb and emit electromagnetic radiation (see Planck, 1900a,
p. 69). The idea that oscillating ions were responsible for the emission and absorption of electro-
magnetic radiation by matter was emphasized, in particular, by Hendrik Antoon Lorentz. In his
book Versuch einer Theorie der electrischen und optischen Erscheinungen in bewegten Korpern
(Attempt at a Theory of the Electric and Optical Phenomenon in Moving Bodies, Lorentz, 1895),
dealing with electrical and optical phenomenon in moving bodies, in which Lorentz presented this
idea, he also drew attention to the fact that his formulae for the components of the electric and
 Max Planck and the Law of Blackbody Radiation 29

the investigation of resonating systems during the following years, including, for
example, the treatment of the damping effect arising from the emission of radiation
(Planck, 1896). He explained the goal of his endeavours in the following words:
The study of conservative damping [i.e., of radiation damping, which Planck called 'conser-
vative,' because it did not violate energy conservation] appears to me to be of fundamental
importance due to the fact that through it one's view is opened towards the possibility
of a general explanation of irreversible processes with the help of conservative forces —
a problem which confronts the theoretical research in physics more urgently every day.
(Planck, 1896, p. 154)
By the 'pressing problem' Planck referred to the question of whether the thermody-
namic concept of entropy could be defined in a rational way either in mechanics or
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

in electrodynamics. Ludwig Boltzmann had proposed an expression for the entropy

derived from kinetic gas theory (Boltzmann, 1872). But in 1896 Planck's assistant
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Ernst Zermelo suggested that Boltzmann's result was not correct (Zermelo, 1896).
He argued, especially, that according to a mechanical theorem of Henri Poincare, a
conservative system of gas molecules would always return to its initial state after
a finite period of time; hence, a mechanical definition of entropy which implied an
evolution of the system in time towards a more probable state in the sense of Boltz-
mann was not at all possible. Against this objection of Zermelo, Boltzmann argued
that his conclusion could be avoided by taking into account the statistical nature of
the H-theorem.18 That is, the entropy could be defined only by a probability con-
sideration, and it was this definition which Planck did not like at all. He preferred
to argue that: 'A rigorous theory of friction on the basis of the kinetic theory [of
matter] will be achieved only with the help of an additional hypothesis' (Planck,
1897a, p. 58). In contrast to this situation in gas theory, however, he hoped that
he would be able to prove (without reference to a probability assumption) the exis-
tence of irreversible processes in a cavity, i.e., a volume surrounded by completely
reflecting walls and filled with (heat) radiation and one resonator a la Hertz, which
absorbs and emits radiation. In particular, he claimed:

Such a resonator will be excited by absorbing energy from the [electromagnetic] radiation
incident upon it from the outside, and it will be damped by emitting energy. Now the
emitted energy will not, in general, be of the same type [especially, it will not have the
same energy distribution] as the absorbed energy; hence the resonator will change by
its vibration the nature of the electromagnetic waves propagating through its vicinity to
some extent. It can be shown that these changes possess, in several respects, a certain

magnetic fields emitted by the oscillating ions agreed with the expressions 'by which Hertz [Hertz,
1888b] has described the oscillations in the neighbourhood of his vibrator' (Lorentz, 1895, p. 54).
Thus, by the end of the century, Planck took into account the identity of Lorentz' molecular
oscillators and his elementary resonators.
18
A discussion about the significance of — H, Boltzmann's mechanical quantity for t h e entropy
of a system, had started already in late 1894, when several British authors, including George
Hartley Bryan (1864-1924) and Samuel Hawksley Burbury (1831-1901), had written critical notes
in Nature on that question. In his reply Boltzmann had pointed out that his Zf-theorem, i.e., the
statement about the decrease of the expression H (and the consequent increase of entropy), could
not be proved by purely mechancial means (Boltzmann, 1895b).
 30 The Golden Age of Theoretical Physics

direction, i.e., the tendency to homogenize [some properties, especially the temperature,
of the incident radiation]. (Planck, 1897a, p. 59)

3. Planck's Work on Irreversible Radiation Processes

In five contributions, entitled ' Uber irreversible Strahlungsvorgange' ('On Irreversi-
ble Radiation Processes') and submitted to the Prussian Academy between February
1897 and May 1899, Planck tried to prove his above statement (Planck, 1897a,b,c,
1898, 1899). These investigations resulted in establishing Wien's law, Eq. (6), as
the law defining the energy distribution of blackbody radiation.
Planck proceeded in the following manner. He first calculated the effect of a
resonator (having a large wavelength and small damping constant) on incident elec-
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

tric waves (Planck, 1897a). After rejecting a criticism of Boltzmann, who claimed
that all processes used by Planck were reversible, Planck went on to prove the ex-
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

istence of irreversibility in a system, in which a resonator situated in the centre of

a spherically shaped totally reflecting cavity absorbs and emits radiation (Planck,
1897c).19 However, in carrying out the proof he had to make an assumption con-
cerning the nature of radiation. He examined this assumption in more detail in
his fourth communication, where he introduced the concept of 'natural radiation'
(natiirliche Strahlung) and claimed: 'It will be shown, in particular, that each radi-
ation process, exhibiting the properties of "natural" radiation, necessarily proceeds
in an irreversible manner, such that the waves after passing the resonator have ra-
diation intensities with smaller fluctuations than before' (Planck, 1898, p. 450). 20
For example, if natural radiation belonging to two different temperatures was ab-
sorbed and reemitted by the resonator, the emitted radiation would correspond to a
more uniform temperature. By invoking the hypothesis of natural radiation Planck
not only succeeded in obtaining a relation between the energy of the resonator and
the intensity of radiation for a given wavelength or frequency, but also in defining
the entropy of radiation by a proper expression such that the change of the total
entropy (of the resonator plus radiation) was always a positive quantity. In the
last communication Planck finally generalized his proof of irreversibility to a cavity
containing heat radiation and many resonators; he further identified the 'electro-
magnetic' entropy with the thermodynamic entropy and derived from it the law of
blackbody radiation (Planck, 1899). He summarized the results of the irreversible
radiation processes in a long paper, which was received by Annalen der Physik on
7 November 1899 (Planck, 1900a).
Within less than three years Planck had achieved his goal of connecting the ther-
modynamic and electrodynamic theories. He had been able, especially, to introduce

In particular, Boltzmann argued that Planck had only used outgoing, but not incoming, spherical
waves in his derivation (Boltzmann, 1897). Planck replied, however, that his assumption about the
incoming wave (that it should always have finite intensity) forbade the use of incoming spherical
waves (Planck, 1897b).
20
Planck defined 'natural radiation' by certain averaging procedures over the phases of the radi-
ation. (See Planck, 1898, p. 468, Eqs. (69).)
 Max Planch and the Law of Blackbody Radiation 31

the concept of irreversible processes in the t r e a t m e n t of systems consisting of radi-

ation a n d resonators absorbing a n d emitting radiation. H e h a d t o p a y a price for
this success; this was the assumption of what he called ' n a t u r a l radiation,' which
could b e interpreted physically as:
If it is stated that an electromagnetic ray possesses the properties of natural radiation,
then in short it should mean the following: The energy of radiation is distributed in a
completely irregular manner among the individual partial vibrations, of which the ray can
be thought to be composed. (Planck, 1900a, p. 73)

In 1899 Planck was aware of the fact t h a t :

The hypothesis of natural radiation, if assumed to be valid for all space points at all times,
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

implies at its core the second law of thermodynamics when applied to the processes of
radiation; that is, it [the hypothesis of natural radiation] is another expression of the same
law. (Planck, 1900a, p. 74) 2 1
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Hence t h e p r o g r a m m e of finding a system in which t h e second law of t h e r m o d y -

namics followed from electrodynamics without further assumptions had failed. T h e
hypothesis of n a t u r a l radiation enabled Planck, however, to define the entropy as a
function of time. Even more:
By identifying the electromagnetic entropy with the thermodynamic entropy, there follows
immediately the electromagnetic definition of the temperature of a ray of heat radiation;
and the maximum of the entropy defines thermal equilibrium, i.e., the state in which the
heat radiation becomes stationary. The law of spectral energy distribution in this state
turns out to be identical with the law which Wien has derived in a different manner, a law
which in some approximation, has been substantiated recently — as is known — by the
experimental investigations of F. Paschen, as well as of O. Lummer and E. Pringsheim.
(Planck, 1900a, p. 74)

4. Planck's Radiation Law

Planck's procedure in deriving the blackbody radiation law involved the following
steps. At first, he established a relation between Uv, the average energy of a res-
onator having frequency v, a n d p„, the energy of incident radiation having t h e same
frequency. T h u s he found from electrodynamics the equation

(7)
""-£?»>•
T h e n he defined t h e entropy of the resonator as

5 = _ ^ m ^ , (8a)
av bv
and t h e entropy of a ray of radiation of frequency v in a given direction as

5=-(^ m ^ + ^ l n ^ V (8b)
\ av bv av bv )
In this connection Planck referred to Boltzmann's hypothesis of 'molecular chaos' ('molecular
Unordnung'), which Boltzmann had introduced in his lectures on gas theory in order to account
for irreversibility in the kinetic theory of matter (Boltzmann, 1895c, p. 21).
 32 The Golden Age of Theoretical Physics

where a and b are constants and Kv and K'v (with pv ~ (4ir/c)(Kv +K'V)) denote the
intensities of the ray in the two main directions of polarization perpendicular to each
other. The sum of the two entropies, given by Eqs. (8a) and (8b), could be shown to
increase in the course of time; hence Planck believed his definitions to make sense.
Finally, by taking the maximum of the total entropy, he obtained a relation between
the resonator energy and the ray intensity in the state of equilibrium, namely,

XJV = \ K V = ^K'V. (9)

In this equilibrium state the temperature could be defined by the relations

1 dS 1 Uv 1 . bv*
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

= = = (10)
T WV -^U) VV^MCV-
Equation (10) implied a relation between Uv and T. In the case of equilibrium,
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Planck derived with the help of Eq. (7) a relation for p„, the energy density of
radiation of frequency v (incident on the resonator), and T, that is,

8nbv3 (11
Pv = — 5 — exp
(-?) • >
Equation (11) thus represented the energy distribution of blackbody radiation and
agreed with Wien's law, given by Eqs. (6), (6a) and (6b), if the quantities v and pv
were rewritten in terms of the wavelength A and the corresponding energy density
distribution p\.22 From Paschen's recent measurements (Paschen, 1899a), Planck
derived the values for his constants a and b, i.e., a = 0.4818 x 1 0 - 1 0 (seconds x
degree Centigrade) and b = 6.885 x 10~ 27 (erg x seconds).
In his paper of November 1899 in Annalen der Physik Planck claimed that
the definition of entropy, Eqs. (8a) and (8b), and therefore also the radiation law,
Eq. (11), were a necessary consequence of the second law of thermodynamics. 'If,
on the other hand, one tries,' he argued, 'to start from any distribution law different
from the one given by Eq. [(H)], and computes from it in turn the entropy, then one
arrives again and again at contradictions with the law of the increase of entropy'
(Planck, 1900a, p. 118). Several months later, in March 1900, he submitted a
further paper on this subject (Planck, 1900b). He had meanwhile convinced himself
that 'the law of the increase of entropy by itself does not suffice to determine the
expression of the entropy as a function of the energy, but that a closer investigation
of the physical significance of the entropy function is necessary for this purpose'
(Planck, 1900b, pp. 730-731). Hence he had to look for another method of defining
the entropies and he did so by presenting a 'method for the direct calculation of the
radiation entropy' (' Weg zur direkten Berechnung der Strahlungsentropie,' Planck,
1900b, p. 720). For this purpose Planck wrote down an expansion for the quantity

^ 2 To see the equality, one has to make use of the equation, J p„ dv = J p\ dX, with both integra-
tions from zero to infinity, and the fact that dv can be written as — (c/A 2 ) dX (c being the velocity
of light in vacuo).
 Max Planck and the Law of Blackbody Radiation 33

dS, expressing the change of the total entropy of the system (cavity radiation plus
resonator), in the vicinity of the equilibrium situation (denoted by the suffix zero),
that is,
<12)
"-"•v-iCs*).-
where dU denotes the change of energy of the resonator associated with the entropy
change dS, and Af7 = U—UQ is the difference of the average energy of the resonator,
U, and the equilibrium energy, Uo. (We have suppressed the suffix v here.) 23 Due
to the equation of motion of the resonator (damped by the emission of radiation),
AU and dU have opposite signs. 24 Since the entropy change, dS, must be a positive
quantity, the factor ^(d2S/dU2)0 on the righthand side of Eq. (12) has to represent
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

a negative quantity, say —f(U), where f(U) is a positive function of the resonator
energy U = Uv. In order to determine f(U), Planck considered n resonators in the
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

cavity and postulated the additivity of the resonator entropies; thus he obtained
from Eq. (12) the result

n dS = -n2 dU AUf(nU). (13)

The obvious means of satisfying this equation was to take f(U) as proportional to
U~x. Then the differential quotient, (d2S/dU2)o, could be written as

fd2S\ _ const.
(14)
[dlp)0- u~
with a positive constant, which depended only on the frequency v of the oscilla-
tor. By integrating Eq. (14), Planck obtained an expression for dS/dU, namely,
—aUln(PU), with factors a and (3 that might depend on the frequency. By com-
paring this expression with Eq. (10), Planck found that a and /3 corresponded to
{au)~l and (bi/)'1, respectively. Then he immediately derived the validity of Wien's
law, Eq. (11), for the energy distribution of blackbody radiation.
At the same time as Planck presented his, as he claimed, rigorous derivation of
Wien's law, the experimentalists reported deviations from the latter. As we have
already mentioned, some members of the Physikalisch- Technische Reichsanstalt had
become increasingly involved in studying the properties of blackbody radiation dur-
ing the 1890s. Interestingly enough, no plan had existed at the Reichsanstalt to
perform measurements to establish the form of Kirchhoff's function. It just so hap-
pened that several scientists working in different departments got into this prob-
lem. Some came to it while investigating certain problems of thermometry, e.g., the

Equation (12) was obtained by expanding both the change of the entropy of the resonator and
of the entropy of the radiation, in a Taylor series and observing that the first-order differential
quotients cancel at the equilibrium position. The subscript u, referring to the frequency of the
oscillator (in Uu), has been dropped in the following derivation.
The time change of the (average) energy of a damped resonator is given by the equation dU/dt+
lav AU = 0, where the positive damping constant <r is related to the emitted intensity of radiation
of frequency v.
 34 The Golden Age of Theoretical Physics

question of determining the temperature in regions where the usual methods (say,
the mercury and gas thermometers) failed. But also a more technical problem, the
search for a suitable standard unit of luminosity, led scientists to develop techniques
which could later be used for the study of blackbody radiation. In particular Otto
Lummer, who had just joined the Reichsanstalt in 1887 and had become the lead-
ing person responsible for optical measurements, turned his attention after 1890 to
the problem of measuring the temperatures of light sources employing incandescent
platinum. 25 To that end he developed both photometric and bolometric methods in
collaboration with his colleague Ferdinand Kurlbaum. 26 In 1895 Lummer published
a short note with Willy Wien, entitled 'Methode zur Priifung des Strahlungsgeset-
zes absolut schwarzer Korper1 ('Method for Testing the Radiation Law of Perfectly
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

Black Bodies'); in it the authors proposed to represent a blackbody by a cavity

which was 'brought to as uniform a temperature as possible, and radiation allowed
to pass outwards through an opening' (Wien and Lummer, 1895, p. 453). 27 A year
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

later Lummer described the method in full detail; he pointed out that all the in-
vestigations carried out hitherto on the properties of blackbody radiation suffered
from errors, because the radiators employed were not completely black, and recom-
mended the use of cavities to determine Kirchhoff's function (Lummer, 1896).
In the first paper on cavity radiation, Wien and Lummer had already suggested
the testing of Stefan-Boltzmann's law and Wien's displacement law by the new
method (Wien and Lummer, 1895). The authors had then started to make prelimi-
nary experiments to achieve this goal. After Wien's departure for Aachen, Lummer
looked for collaborators, and he received the help of Ferdinand Kurlbaum and Ernst
25
O t t o Richard Lummer was born in Gera, Thuringen, on 17 July 1860. He studied at the
University of Berlin, where he obtained his doctorate in 1884 with a thesis on a new interference
effect observed in plane parallel glass plates. He continued to work as an assistant to Hermann
von Helmholtz at the University of Berlin; when Helmholtz became President of the Reichsanstalt,
he took Lummer along with him. Lummer served at the Reichsanstalt first as an assistant, then
(from 1889) as a scientific member, and received t h e title of 'professor' at the same institution in
1894. In 1901 he became Privatdozent at the University of Berlin, and three years later he was
appointed Ordinarius of physics at the University of Breslau. He died in Berslau on 5 July 1925.
Lummer made important contributions to optics and photometry (Lummer's interferometer,
Lummer-Brodhun photometer). He also participated in the discovery of the liquefaction of carbon.
Ferdinand Kurlbaum was born in Burg near Magdeburg on 4 October 1857. He studied at the
Universities of Heidelberg and Berlin and received his doctorate from the latter university in 1887
with a thesis on a new determination of wavelengths of absorption lines in the solar spectrum.
He first went to the Technische Hochschule in Hanover as an assistant with his thesis supervisor
Heinrich Kayser for four years. In 1891 he returned to Berlin and joined the optical laboratory of
the Reichsanstalt, which was directed by Lummer. Ten years later he had his own laboratory at
the Reichsanstalt, and in 1904 he accepted a call to a professorship of physics at the Technische
Hochschule of Berlin, from where he retired in 1925. Kurlbaum died on 29 July 1927 in Berlin.
27
T h e idea of using a cavity for representing a blackbody goes back to Kirchhoff (1860). However,
Wien and Lummer were not the first to think of using Kirchhoff's proposal for experimental
purposes. Already the Danish physician Christian Christiansen (1843-1917), one of the discoverers
of the anomalous dispersion of light, had experimented with the absorption of (visible) radiation
by cavities having a small hole and observed that they could be taken as a good approximation of
a blackbody (Christiansen, 1884). A few months later, independently of Wien and Lummer, the
American physicist Edward St. John, then a student at the University of Berlin, also noticed the
blackbody properties of a cavity radiator (St. John, 1895).
 Max Planck and the Law of Blackbody Radiation 35

Pringsheim for his work on blackbody radiation. 28 In less than three years these
three experimentalists improved the techniques of observation to such an extent
that the problem of the measurement of Kirchhoff's function could be attacked
in earnest. 29 In performing their investigation, however, Lummer and Pringsheim
made use of another experimental development which concerned the analysis of
very long wavelengths. The principal contribution in that field had been made by
Heinrich Rubens, since 1892 Privatdozent at the University of Berlin and from 1895
Professor of Physics at the Technische Hochschule in Berlin. 30 In 1889 he had al-
ready begun to measure the wavelengths of invisible infrared radiation with the help
of a Rowland grating and the bolometer method. During the 1890s he penetrated
further into the infrared spectral region by using various techniques. Finally, in
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

late 1896 he published, together with a visitor from the United States, Ernest Fox
Nichols (1869-1924), a new method for measuring long wavelengths (Rubens and
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Nichols, 1896). This new method, which was later called the method of 'residual
rays' ('Reststrahlenmethode'), made use of the fact that all substances reflect radi-
ation especially strongly in the region of strong absorption, hence it was possible
to isolate certain wavelengths by multiple reflection; thus, by 1898, wavelengths of
61.1 /i were reached by using a sylvine crystal as the reflecting substance (Rubens
and Aschkinass, 1898). The tools were thus made ready for a fresh attack on the
empirical determination of the radiation law.
On 3 February 1899, at a session of the German Physical Society in Berlin,
Lummer presented the first results of his and Pringsheim's investigations on the en-
ergy distribution in the spectrum of blackbody radiation (Lummer and Pringsheim,

28
E r n s t Pringsheim was born in Breslau on 11 July 1859. He studied physics and mathematics in
Heidelberg, Breslau, and Berlin, from 1877 to 1882. He received his doctorate at the University of
Berlin in 1882, where he also became Privatdozent four years later and Titular Professor in 1896.
In 1905 he accepted a call to a full professorship in theoretical physics at the University of Breslau,
where he joined Otto Lummer, his collaborator for many years. Pringsheim worked on problems
of radiation including luminescence, on optics (interferometry) and on physics of the sun. He died
on 28 June 1917 in Breslau.
29
Special progress in that direction was achieved through the construction of a suitable cavity,
namely, an electrically heated platinum cylinder which was blackened inside with iron oxide and
divided by diaphragms and enclosed in a larger asbestos cylinder (Lummer and Kurlbaum, 1898).
30
Heinrich Leopold Rubens was born in Wiesbaden on 30 March 1865. He first studied electri-
cal engineering at the Technical Universities (Technische Hochschulen) of Darmstadt and Berlin
(1884-1885), then physics at the Universities of Strasbourg and Berlin. He obtained his doctorate
under August Kundt from the University of Berlin in 1889; he stayed on there as an assistant and
Privatdozent until he was appointed in 1896 an 'aufierplanmafiiger' professor of physics at the
Technische Hochschule of Berlin. He became a full professor in 1900 and moved six years later to
the University of Berlin, succeeding Paul Drude. He devoted nearly his entire scientific career to
the study of electromagnetic waves, especially the infrared region which he extended immensely
by his researches. He received many honours, including t h e Rumford Medal of the Royal Society
and an honorary D.Sc. from Cambridge University; he was a member of the scientific academies
of Berlin and Gottingen. Rubens died in Berlin on 17 July 1922. The importance of his role
in quantum theory was later described by Max Planck as follows: 'Without the intervention of
Rubens the formulation of the radiation law and thereby the foundation of quantum theory would
perhaps have [arisen] in quite a different manner, or perhaps not have developed in Germany at
all' (Planck, 1923a, p. cxi).
 36 The Golden Age of Theoretical Physics

1899a). They had observed wavelengths between 0.2 (j, and 6 fi emitted from cav-
ities heated to temperatures between 800 and 1400 degrees absolute; they found
that the energy distribution agreed in general with Paschen's law, Eq. (2), although
the constant c^ seemed to increase systematically with decreasing temperatures. In
order to eliminate experimental errors Lummer and Pringsheim continued to im-
prove their methods in the following months. Later in the year, on 3 November
1899, Lummer again spoke at a session of the German Physical Society; now he
was certain that there were 'discrepancies of a systematic nature between theory
and experiment' ('Abweichungen zwischen Theorie und Experiment systematischer
Natur,' Lummer and Pringsheim, 1899b, p. 222) which could not be attributed to
experimental errors. In contrast to Lummer and Pringsheim, Friedrich Paschen —
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

in two contributions to the Prussian Academy, which were presented by Planck at

the sessions of 27 April and 7 December 1899, respectively — did not find any devi-
ations from Wien's law (Paschen, 1899a,b), and the situation remained undecided
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

in the beginning of 1900.

During 1900 the meetings of the German Physical Society in Berlin became the
main forum where the problems connected with the law of blackbody radiation were
discussed most frequently because most of the leading experts were resident mem-
bers; besides Max Planck, Otto Lummer, Ernst Pringsheim, Ferdinand Kurlbaum
and Heinrich Rubens, all of whom we have already mentioned, there were scientists
like Ludwig Holborn, Eugen Jahnke, Friedrich Kohlrausch (the Acting President
of the Reichsanstalt), Max Ferdinand Thiesen and Emil Warburg, who became in-
terested in the subject. 31 The first session on the problems of blackbody radiation
took place on 2 February. In it Max Thiesen presented a theoretical discussion of
blackbody radiation (which he called 'black radiation'), Ernst Pringsheim reported
about new measurements by Lummer and himself, and finally Planck discussed the
assumptions which had entered into his derivation of Wien's law (Planck, 1900b). 32
Thiesen analyzed the structure of the existing laws of blackbody radiation — i.e.,
the Stefan-Boltzmann law as well as Wien's displacement law and the radiation law,

31
T h e German Physical Society or, more accurately, the Deutsche Physikalische Gesellschaft zu
Berlin, grew out of the Berliner Physikalische Gesellschaft which was founded on 14 January 1845
after a colloquium by Gustav Magnus at the University of Berlin. Among the founders one finds
the names of the chemist Wilhelm Heintz, the physiologists Emil du Bois Reymond and Ernst
Brticke, and the physicists Hermann Knoblauch, Wilhelm Beetz and Gustav Karsten; they were
soon joined by Hermann von Helmholtz and Werner Siemens (later 'von' Siemens). The sessions
of the Gesellschaft took place regularly in the Institute of Physics of the University of Berlin;
towards the end of the century there took place about one session every two weeks, except during
the semester vacations. At the end of 1898 the members of the Berlin Physical Society decided to
change the name to German Physical Society; the members were partly Berlin residents (about
50%) and partly nonresidents. (For historical details, see Schwalbe, 1900.)
32
T h e report of this session, after mentioning the talks of Thiesen and Pringsheim, stated:
Following the two talks, Mr. Planck discusses in detail the assumptions on which the
radiation theory developed by him is founded, and reports on a direct 'Deduktion der
Strahlungs-Entropie aus dem zweiten Hauptsatz der Thermodynamik' ('Deduction of the
Radiation Entropy from the Second Law of Thermodynamics'), which he found recently
and will publish soon. (Verh. d. Deutsch. Phys. Ges. (2) 2, Nr. 3, 1900, p. 37)
 Max Planck and the Law of Blackbody Radiation 37

Eq. (6) — and suggested a more complicated formula, generalizing Wien's Eq. (6)
in order to account for the deviations observed especially by Lummer and Pring-
sheim (Thiesen, 1900a). Pringsheim showed that these deviations became more
significant if one studied — for the same temperature of the emitter — the long
wavelengths of the spectrum. (The new experiments went up to 18 ji: Lummer and
Pringsheim, 1900).33
The February session of the German Physical Society started a series of consid-
erations to examine the validity of Wien's radiation law. Thus Lummer, together
with Eugen Jahnke, a high school teacher, submitted a paper to Annalen der Physik
in which they made proposals to generalize Eq. (6) (Lummer and Jahnke, 1900).
Such attacks on his equation encouraged Willy Wien to reply to the criticisms; he
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

did so in his talk at the International Congress of Physics (Congres International

de Physique), held at Paris on the occasion of the World's Fair from 6 to 12 Au-
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

gust 1900, and in a paper submitted to Annalen der Physik in October 1900 (Wien,
1900a and 1900b, respectively). He came to the conclusion that on the basis of ther-
modynamic and kinetic-theoretic arguments it followed that Eq. (6) must describe
the energy distribution among short but not among long wavelengths.
Wien's conclusion was confirmed by the results obtained by Heinrich Rubens
and Ferdinand Kurlbaum, who investigated the spectrum of cavity radiation up to
the longest wavelengths obtained until then, namely, the residual rays of rock-salt
having a wavelength of 51.2 fi; actually, the crucial quantity was the product of
wavelength and temperature, which reached values up to 90,000 in the measure-
ments of Rubens and Kurlbaum, in contrast to the value of 32,000 obtained earlier
by Lummer and Pringsheim (Rubens and Kurlbaum, 1900). The main outcome of
the observations of Rubens and Kurlbaum could be stated briefly as follows: for
very long wavelengths A and very high temperatures T, the intensity of radiation
increases in proportion to T, i.e.,

px = const. T, (15)
in agreement with the formulae derived earlier by Lord Rayleigh (1900b) and Lum-
mer and Jahnke (1900). (See Rubens and Kurlbaum, 1900, p. 941.)
Planck received the news about the results of Rubens and Kurlbaum before
their public announcement. In a report about these events Planck's student Gerhard
Hettner recalled later: 'When on Sunday, 7 October 1900, Rubens together with his
wife visited Planck, the discussion turned to the measurements with which Rubens
was occupied. He [Rubens] said that for the longest wavelengths [which he could
achieve], the law recently proposed by Lord Rayleigh was valid' (Hettner, 1922,
p. 1036). 34 On receiving this information from Rubens, Planck set down and studied
^Pringsheim's report was not published in spring 1900 but only in the fall. The publication in the
Verhandlungen represented essentially the content of a talk given on 18 September 1900 before
the Versammlung Deutscher Naturforscher under Arzte in Aachen (Lummer and Pringsheim,
1900, p. 163).
34
Hettner obtained the information from Planck himself. (See Kangro, 1970. Section 8.8; English
edition, pp. 198-200.)
 38 The Golden Age of Theoretical Physics

the theoretical implications for the equilibrium entropy; he noticed especially that
if Rubens' results were correct, then for long wavelengths — or, more accurately,
for large XT — the entropy had to satisfy the equation 35

fd2S\ _ const.
(16)

Then he combined Eq. (16) with Eq. (14), which was valid for short wavelengths
(more accurately, small AT) and obtained the following equation for the equilibrium
entropy:
(d2S\ a
{U)
\dU*)0 Utf + U)'
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

where a and /? were constants depending on the wavelengths of the radiation. Planck
then integrated Eq. (17) to obtain the expression for (dS/dU)o, which he put equal
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

to the inverse of the absolute temperature. By again integrating the relation be-
tween entropy and temperature, he found an equation for the energy of the res-
onator, from which he derived — with the help of Eq. (7) — a new radiation
formula, that is,

where cj and c<± were constants similar to the ones entering Paschen's law, Eq. (2). 36
As Hettner reported later on these events: 'The same evening still he [Planck] re-
ported this formula to Rubens on a postcard, which the latter received the following
morning . . . One or two days later Rubens again went to Planck, and was able to
bring him the news that the new formula agreed perfectly with his [Rubens'] obser-
vations' (Hettner, 1922, p. 1036). At the following meeting of the German Physical
Society, on Friday, 19 October 1900, Kurlbaum talked about his and Ruben's exper-
iments on 'the emission of long wavelengths from black bodies' (' Uber die Emission
Langer Wellen durch den Schwarzen Korper1), The report on this meeting, given
35
E q u a t i o n (16) follows from the definition of the entropy, (dS/dU)o = l/T, and the proportional-
ity of the energy U and the temperature T. (This gives the relation (dS/dU)o ~ 1/U, from which
one derives the expression for (d?S/dU2)o. The index zero on the left-hand side of Eq. (16) is
added to emphasize the fact t h a t this equation describes only the equilibrium situation. Again, the
subscripts referring to frequency and wavelength have been dropped in the following derivations.)
36
E q u a t i o n (18) was obtained as follows. From the integration of Eq. (17), one found an expression
for (dS/dU)o, which could be put equal to the inverse of the absolute temperature, i.e.,

( du)0~ P " 0 + U ~ T'

By integrating this equation, Planck found the result

"-'M-s?)-1]-1
for the equilibrium energy of the resonator of a given frequency, U = UV{T). The result agreed
with Eq. (18) if he took into account Eq. (7) between Uv and pv and rewrote it as an equation
for px, t h e energy distribution over the wavelengths. The constants a and /3 of Eq. (17) were now
replaced by the constants ci and C2, that is, a = —c\/c2C2 and f3 = c\/c2x\.
 Max Planck and the Law of Blackbody Radiation 39

in the Verhandlungen t h e n stated: 'In the lively discussion following this talk,
Mr. M. Planck spoke on "An Improvement of Wien's Spectral L a w " . ' 3 7
W i t h Planck's formula, whose validity for all wavelengths a n d all t e m p e r a t u r e s
became established again a n d again in the following years, the correct law for the
energy distribution of blackbody radiation had been given. 3 8 For Planck, however,
the immediate a n d more fundamental t a s k was t o provide t h e theoretical foundation
and physical motivation for the basic Eq. (17). T h i s task led him t o the discovery
of a new constant of nature: the q u a n t u m of action.

5. T h e S i g n i f i c a n c e o f t h e C o n s t a n t s i n P l a n c k ' s L a w
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

In October 1900 it had become certain on experimental grounds t h a t Wien's law,

Eq. (6), did not describe the entire energy distribution in blackbody radiation,
and M a x Planck — so far a firm believer in Wein's law — had proposed a new law,
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

Eq. (18), which agreed quantitatively with the most recent results (Planck, 1900c). 3 9
T h e new situation implied t h a t some of t h e assumptions, which P l a n c k h a d m a d e
earlier in his theory of blackbody radiation, were definitely wrong. Planck h a d
immediately identified the crucial point on which he h a d t o improve: while he h a d
previously assumed t h a t (d?S/dU2)o, t h e second derivative of the resonator entropy
at equilibrium, was proportional to U~x, the inverse of the average resonator energy,
he now arrived at the conclusion t h a t Eq. (14) should be given u p . I n particular,
he had remarked:

I believe, on the contrary, that it is possible — though still not easy to grasp and in any case
difficult to prove — that the expression on the left-hand side [of Eq. (14] does not possess
in general the significance I had attributed to it. In other words: the values assumed by
£/„, dUn and AZ7„ [denoting, respectively, the total energy of n resonators in a cavity and
its changes] do not suffice at all for determining the change of entropy [of the n resonators]
considered; in addition, U [i.e., the energies of the individual resonators] must be known.
By following this idea I have finally been able to construct totally arbitrary expressions for
the entropy, which — though more complicated than Wien's expression [i.e., the entropy
derived from Wien's law and given by Eq. (8a)] — still seem to satisfy, as completely
as the latter, all requirements of thermodynamics and electrodynamics. (Planck, 1900c,
p. 203)

T h u s he also studied the entropy function S satisfying Eq. (17), t h a t is,

*—[(£ +i ) h (£ +i )-£ ta & (19)

37
See Verh. d. Deutsch. Phys. Ges. 2 (1900), p. 181.
38
In November 1900 Paschen submitted a paper to Annalen der Physik in which he reported that
his own recent measurements agreed with those of Rubens and Kurlbaum (Paschen, 1901).
39
Not only did the measurements of the long wavelength distribution of Rubens and Kurlbaum
(1900), quoted above, contribute to Planck's conversion, but Friedrich Paschen, whose measure-
ments had always confirmed Eq. (6) (e.g., Paschen, 1899a,b), had also written a letter to Planck at
about that time pointing to 'appreciable departures from Wien's law.' (See Planck, 1900c, p. 202,
footnote 1.)
 40 The Golden Age of Theoretical Physics

w i t h t h e constant /3' = (3/i> a n d the suffix zero denoting t h e equilibrium function. 4 0

T h i s entropy h a d t h e n yielded the radiation law, Eq. (18).
Now, after he had proposed the successful formula, Planck was faced with the
task of obtaining a theoretical derivation of t h e expression for t h e equilibrium en-
tropy So- He h a d already observed earlier — in February 1900 — t h a t t h e r m o d y -
namic a r g u m e n t s alone did not suffice to define t h e entropy function of t h e resonator
and, therefore, also of t h e blackbody radiation (Planck, 1900b). T h e additional ar-
gument which h a d enabled him to arrive at, as he believed, a unique definition of
the entropy via Eq. (14), h a d now been discovered t o fail. O n what basis could he,
in t h e light of these negative experiences, claim t h a t Eq. (18) yielded, u p to addi-
tional constants, the correct entropy of t h e resonator in a cavity? How could he
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

justify, in particular, t h e existence of two constants, a a n d /3', a n d why did he not

have t o introduce any other constants? Would h e have t o give u p essential p a r t s of
his theory of irreversible radiation processes which he h a d developed over so m a n y
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

years? After all, Wien, in his paper in Annalen der Physik, h a d just recently raised
two a r g u m e n t s against this theory, namely:

First, one misses the demonstration that the hypothesis of natural radiation which has
been introduced [by Planck] is the only one that might lead to irreversibility. Hence it
remains doubtful whether the processes considered [i.e., the absorption and reemission of
radiation by the resonators in the cavity] are related in any way to the heat radiation.
Secondly, in the derivation of the expression for the entropy the radiating resonators are
assumed to be independent of each other. But the derivation [of the expression for the
entropy, Eq. (14)] depends on the assumption that several resonators exist. In my opinion,
this seems to imply a contradiction. (Wien, 1900b, pp. 538-539)

T h o u g h Planck accepted t h e opinion t h a t this theory was, so far, incomplete,

he did not for a m o m e n t think of giving u p his original plan of research, namely,
the principal idea t h a t t h e condition of irreversibility of radiation processes would
determine t h e law of blackbody radiation. He also disagreed strongly with Wien on
another point of principle, which the latter h a d stated in his p a p e r in Annalen der
Physik as follows:
I must emphasize, first, that I still stick, in contrast to Mr. Planck [see Planck, 1900a,
p. 725], to my previously stated opinion [see Wien, 1893b, p. 633] that short and long
electromagnetic waves — as far as their relations to heat radiation are concerned — differ
more than just quantitatively from each other. With respect to the process of absorption
it is assumed generally that longer waves can be described by a single vector or, what
amounts to the same, that matter can be regarded in that case as being continuous;
however, for shorter wavelengths the influence of the molecular constitution of bodies

40
In the integration of Eq. (17) we have suppressed all terms not depending on the (average)
energy of the resonator U. Note that in Eq. (19), U is a function of the frequency v and the
temperature T.
41
Planck had listened to Wien's arguments already at the Paris Conference (Wien, 1900a). He
read them again in Wien's paper for Annalen der Physik before its publication because he was
assisting the editor of the journal (Wien, 1900b). Planck held seen the paper certainly soon after
it was received on 12 October, for he replied to some of the other questions raised in Wien's paper
in November (Planck, 1900d,e).
 Max Planck and the Law of Blackbody Radiation 41

comes into play. Exactly the same must be true for emission processes. Therefore, I
believe it to be improbable from the very beginning to assume that a radiation law, which
rests on the molecular hypothesis, should also be valid for very long waves. The argument
[of the radiation law] with the empirical data in the case of short wavelengths obviously
shows that the assumptions made [in the derivation of the law] are approximately satisfied
for waves whose wavelengths are not too long. (Wien, 1900b, pp. 537-538)

If a qualitative difference existed, as Wien claimed, between the heat radiations

of long and short wavelengths, then the unity of electromagnetic radiation had to
be given up, and the great advance brought about by Maxwell's electromagnetic
theory would be lost again. Such an attitude contradicted the scientific goals of
Max Planck, who had always sought to establish relations and connections between
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

different phenomenon rather than to separate them. He was convinced that a

unified theory had to describe all blackbody radiation and that this theory was
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

based on electrodynamics and thermodynamics. On the other hand, he recognized

the necessity of taking into account even other physical ideas in order to derive a
radiation formula applicable to all wavelengths.
As to the direction in which to look for these supplementary ideas, Planck could
obtain some indication from his previous experience in the theory of cavity radiation
as well as from Wien's critical remarks quoted above. In order to explain Eq. (17),
the Ansatz for the entropy function, it was evidently necessary to consider several
resonators in the cavity and to assume that the entropy would not only depend on
the total energy of all resonators but also on the distribution of energy among the
individual resonators. Still, the question remained open as to how one might pro-
ceed in detail. The most obvious method to apply seemed to follow Wien's example
of 1896, that is, to treat the objects emitting heat radiation in a way similar to the
molecules in kinetic theory. However, if one wanted to arrive at a theory including
the long waves as well, more features of kinetic theory had to be used than Wien had
employed earlier. Now the most detailed and complete scheme of kinetic theory had
been developed by Ludwig Boltzmann. Planck had been aware of Boltzmann's work
for quite some time. He had even had the opportunity of discussing with Boltzmann
a specific question of kinetic theory, namely, the derivation of the Maxwellian veloc-
ity distribution for molecules in a gas. 42 However, Planck — though he had joined
Boltzmann in fighting against the views of the energeticists — had not developed
closer relations with him. As he explained later: 'In particular he [Boltzmann] was
unhappy about the fact that I was not only indifferent towards the atomistic theory,
which constituted the foundation of his entire research work, but even opposed it'
(Planck, 1948a, p. 21). In particular, he had criticized the probabilistic interpre-
tation of entropy. However, in the course of his treatment of the theory of heat
radiation, Planck had withdrawn his earlier criticism and increasingly approached

42
Planck had edited Gustav Kirchhoff's lectures on the theory of heat (Kirchhoff, 1894), and Boltz-
mann had criticized a passage in the book dealing with a proof of Maxwell's velocity distribution
formula (Boltzmann, 1894). Planck had then replied to Boltzmann's objections and formulated
the proof more accurately (Planck, 1894c), with which Boltzmann agreed (Boltzmann, 1895a).
 42 The Golden Age of Theoretical Physics

Boltzmann's views on the origin of irreversibility, especially by introducing the hy-

pothesis of 'natural radiation' and recognizing the similarity of this concept with
that of 'molecular chaos,' which Boltzmann had introduced in kinetic theory. Thus
in fall 1900 Planck was prepared well enough and willing to search in Boltzmann's
statistical theory of matter for the physical interpretation of his radiation formula,
Eq. (18).
This problem [he recalled later in his Nobel lecture] led me automatically to a consideration
of the connection between entropy and probability, that is, to Boltzmann's trends of ideas,
until after some weeks of the most strenuous work of my life, light came into the darkness
and a new, hitherto undreamt of perspective began to open up for me (Planck, 1920,
p. 5)
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

The reconstruction of the step which led Planck to the interpretation of his radi-
ation formula, Eq. (18), must begin by analyzing the expression for the equilibrium
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

entropy of the resonator on the right-hand side of Eq. (19). For some time Planck
tried to use only thermodynamic and electrodynamic arguments, but without any
success. Then, 'since no other path [seemed to be] open,' he turned to the 'method
of Boltzmann' (Planck, 1943, p. 157). This method employed the statistical inter-
pretation of the entropy, in particular the fact that the entropy of any molecular
system in a given state could be identified with the 'permutation measure' ('Permu-
tationsmafi'), that is, the logarithm of the number of 'complexions' or possibilities
of permuting the molecules without changing the state of the system (Boltzmann,
1877). Planck now formulated Boltzmann's relation as

S = k\nV, (20)

where V denotes the number of complexions and A; is a natural constant. With

Eq. (20) in mind he then returned to the interpretation of So, the equilibrium en-
tropy of the resonator given by Eq. (19). The expression within the square brackets
on the right-hand side of Eq. (19) should denote the permutation measure of the
resonators, but it was not immediately obvious how. Surely, a permutation measure
made sense only in the case of a system of, say, n resonators, in which case it de-
scribed the distribution of energy among the individual resonators. Would Planck
be able to find in Boltzmann's work an expression for the permutation measure of
a system of n objects having a structure similar to the expression within the square
brackets on the right-hand side of Eq. (19)? For this purpose Planck studied in de-
tail Boltzmann's great memoir, and he discovered right away the desired expression
in its first section (Boltzmann, 1877). There Boltzmann had given an equation for
J , the total number of complexions or possibilities of distributing A discrete, equal
energy values e among n molecules, as

(A + n - l ) l
J (
A!(n-1)! • ^>
 Max Planck and the Law of Blackbody Radiation 43

He had rewritten this equation for large values of n and A with the help of Stirling's
formula as
1 (A+ n - ! ) * + " - ' / »
V^("-l)n-1/2AA+1/2' { }

Then the logarithm of J, after neglecting the terms —1 and ± 5 in comparison to n

and A, would take the form

A V /A \ A, A
In J = n - + l l n - + l - - l n - (22)
n I \n I n n

(Note that on the right-hand side of this equation a subtractive term of magnitude
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

ln\/27r has been dropped.) Planck immediately noticed the similarity between In J
and his expression, Eq. (19), for the equilibrium entropy of the resonator in the
following way: if he considered n resonators in a cavity having the equilibrium
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

entropy nSo, then the expressions within the square brackets of Eqs. (19) and (22)
were identical provided he put \/n equal to U/P'v. Now U, the average energy
of one resonator, assumed the value (A/n)e; hence he concluded that the discrete
energy value in blackbody radiation was given by the equation

e = 0v. (23)

Thus the constant /3', which Planck would later denote by the letter h, possessed
the dimension of an action. It determined, when multiplied by the frequency v, the
size of the energy elements. Finally, in order to make the agreement between So
and kin J complete, Planck identified the constant —a with the natural constant k
occurring in Eq. (20).
The question arises whether Planck was really allowed to take over Boltzmann's
expression for J to describe the equilibrium situation of his cavity resonators. Boltz-
mann, in his memoir of 1877, had given a different prescription for calculating the
equilibrium entropy of molecular systems. He had first talked about the distri-
bution of A energy values e among n molecules and had presented the method of
obtaining the number of complexions; then he had claimed that the equilibrium
state of the system was defined by taking the maximum of all possible numbers of
complexions. Planck did not follow this procedure at all, but just referred to the
expression J, which entered into Boltzmann's calculations as the factor normalizing
V, the number of complexions for any given state, such that the quotient (V/J)
could be interpreted as a probability. More than thirty years later Planck tried to
recall the psychological reasons that motivated his step:
In short [he wrote to Robert Williams Wood], I can characterize the whole procedure as
an act of despair, since, by nature I am peaceable and opposed to doubtful adventures.
However, I had already fought for 6 years (since 1894) with the problem of equilibrium
between radiation and matter without arriving at any successful result. I was aware that
this problem was of fundamental importance in physics, and I knew the formula describing
the energy distribution in the normal spectrum [i.e., the spectrum of a blackbody]; hence a
 44 The Golden Age of Theoretical Physics

t h e o r e t i c a l i n t e r p r e t a t i o n had t o b e found a t a n y price, however high it m i g h t b e . (Planck

t o W o o d , 7 O c t o b e r 1931)

In that desperate situation Planck would have turned to any suitable formula of
Boltzmann's which had to do with the number of complexions. Actually, the choice
of taking J really agreed with Planck's goals; only one did not have to stick to its
interpretation in Boltzmann's original memoir. Planck was interested in calculating
the entropy of the equilibrium state; thus he interpreted J, Eq. (21), right away
as the number of complexions for the equilibrium state and, consistent with it, he
identified the expression (A/n)e with the average energy of the resonators in the
equilibrium state. At that moment he was not interested in Boltzmann's procedure
for obtaining the equilibrium distribution by selecting from different distributions
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

the one associated with the maximum number of complexions; he rather assumed
that he knew the equilibrium state already. And, after all, that was indeed the case
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

in his considerations, which were aimed at obtaining a physical interpretation of

the entropy SQ derived from the radiation law, Eq. (18), i.e., from the equilibrium
situation. 43
Planck had the physical interpretation of his radiation law in hand before the
middle of November 1900, but he presented his results first at the meeting of the
German Physical Society in Berlin on 14 December 1900 in a contribution entitled
'Zur Theorie des Gesetzes der Energieverteilung im Normalspektrum' ('On the The-
ory of Law of Energy Distribution in Normal Spectrum,' Planck, 1900f). 44 In this
contribution he outlined in short 'the most essential point of the whole theory as
clearly as possible' (Planck, 1900f, p. 238). Thus, after emphasizing the necessity
of introducing Boltzmann's probability arguments in the theory of blackbody radi-
ation, he started right away by considering the distribution of a given amount of
energy, E, among N cavity resonators having the frequency v. He noted:
If E is considered as an infinitely divisible quantity, the distribution can be made in an
infinite number of ways. However, we consider — and this is the most important point of
the entire calculation — E as being composed of a completely definite number of finite,
equal parts, and make use for that purpose of the natural constant h = 6.55 x 10 - 2 7 [erg •
sec]. This constant, when multiplied by the common frequency v of the resonators, yields
the energy element e in ergs; and by dividing E by e we obtain P, the number of energy
elements which have to be distributed among the N resonators. If the quotient [E/e] thus
calculated does not happen to be an integral number, then one has to take for P an integer
close to it [the quotient]. (Planck, 1900f, pp. 239-240)

He then calculated 11, the number of complexions for the distribution of the en-
ergy E among the N resonators of frequency u, according to the combinatorial
equation (20), Further, he considered the cavity to contain also resonators of dif-
43
For the reasons given above, we disagree with Thomas Kuhn's conclusion (Kuhn, 1978, pp. 282-
283, footnote 20) t h a t Planck had to take his formula for the number of complexions in radiation
theory from sources different from Boltzmann's 1877 memoir.
44
Concerning the status of his theoretical derivation of the radiation law, Planck wrote to Wien
on 13 November 1900: 'My new formula [Eq. (18)] is well satisfied; I now have also obtained a
theory for it, which I shall present in four weeks at the Physical Society here [in Berlin].'
 Max Planck and the Law of Blackbody Radiation 45

ferent frequencies, say N' of frequency v',N" of frequency i>", etc., among which
the energy amounts of E',E"', etc., were distributed in 1Z',7Z", etc., ways. Evi-
dently, the number of complexions denoting the possibilities of the distribution of
discrete energy elements e (= hu), e' (= hu'), e" (= his"), etc., among the resonators
was identical with the product 72. • 72.' - 1Z" • • •. In case of thermal equilibrium, a
temperature could be defined through the equation
1 _ dlnTZo
(24)
T~ dE0
where 1ZQ is the maximum value of the number of complexions TZ-TZ' • 1Z" •• • (or
the number of complexions in equilibrium) and EQ (= E + E' + E" + • • •) denotes
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

the sum of all energies distributed among the frequencies (u, u', u",...) of the cavity
resonators. Planck did not carry out any maximalization procedure with the total
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

number of complexions; he rather assumed that the equilibrium distributions were

already established. Then it was sufficient to study the (equilibrium) energy distri-
bution among the resonators of a single frequency, say u, and to obtain from it the
expression for C/„, the average energy of the resonator (at temperature T), that is,

U„ = UV{T) = hu exp (25)

(S
When Planck combined Eq. (25) with Eq. (7) he obtained an equation for the energy
distribution among the frequencies in blackbody radiation, which he could easily
transform into Eq. (18) for the distribution among the wavelengths.
In his Atombau und Spektrallinien, Arnold Sommerfeld called 14 December 1900
the 'birthday of quantum theory' (' Geburtstag der Quantentheorie,'' Sommerfeld,
1919, p. 213). He referred in particular to the fact which Planck had considered to be
the 'most essential point' of his derivation of the blackbody radiation law: namely,
the assumption that the energy is distributed among the cavity resonators only
in integral multiples of finite energy elements. The existence of these finite energy
elements followed from the application of Boltzmann's relation between entropy and
number of complexions also to the system of resonators in a cavity, i.e., from the
equation
S = kln1Z. (26)
Moreover, the application of Eq. (26) to the theory of blackbody radiation led to
an important consequence, which Planck also mentioned in his talk at the German
Physical Society in Berlin on 14 December 1900. If one considered a system of gas
molecules and of radiating resonators in equilibrium, the total entropy was given
by the sum of the entropies of the molecules and the resonators, that is,

S0 = f ln(P0TZo) = f\nVo + f\n1ZQ, (27)

where Vo and 1ZQ denote the respective numbers of complexions. Now Planck
observed that the constant factor / had to be proportional to the universal gas
 46 The Golden Age of Theoretical Physics

constant R, with a proportionality factor u; denoting the ratio of the mass of the
molecule to the mass of a gram-atom (or gram-molecule) of the gas (Boltzmann,
1877). Evidently, the inverse of u then represented the number of molecules in
one gram-atom, the so-called Avogadro number (or, in German scientific literature,
Loschmidt number) NQ. Since, due to Eqs. (20) and (27), / must be equal to the
constnat k of radiation theory — k may be expressed in terms of the constants
c\ and ci of the blackbody radiation formula, Eq. (18), as k = ci/c2C2 — Planck
obtained for No the equation

JV0-1=W = ; | , (28)
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

Upon substituting the observed values for k and R (i.e., k = 1.346 x 1 0 - 1 6 and
R = 8.31 x 10 7 erg/°C): 'A real molecule is 1.62 x 10 _ 2 4 -th fraction of a gram-
molecule, or: an atom of hydrogen weighs 1.64 x 1 0 - 2 4 gm, since H [i.e., the weight
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

of one gram-atom of hydrogen in units of grams] = 1.01; or, a gram molecule of any
substance consists of 1/w = 6.175 x 10 23 real molecules' (Planck, 1900f, p. 244).
He noted that this value of UJ~X or No agreed well with the best estimate known
from kinetic gas theory, namely, N0 = 6.40 x 10 23 (Meyer, 1899, p. 337). With the
help of the ratio k/R he also calculated the 'elementary quantum of electricity e,
that is, the electric charge of a positive single-valued ion or electron' according to
the equation
e = F-^, (29)

where F (Faraday's constant) is the charge of a gram-atom of single-valued ions.

Thus he found that the unit charge assumed the value 4.69 x 1 0 - 1 0 electrostatic
units, adding that 'F. Richarz [1894] found 1.29 x 1 0 - 1 0 , while J.J. Thomson [1898b]
has recently found 6.5 x K T 1 0 ' (Planck, 1900f, p. 245).
Planck claimed repeatedly that the relations between the constants of the radi-
ation law, c\ (= c3h) and Ci (— ch/k), and the constants LJ and e which determine
the 'elementary quanta of matter and electricity' had to hold as accurately as the
given experimental data.
Hence the accuracy of the numbers calculated [i.e., of 6.175 x 1023 for ui~l and of 4.69 x
10" 10 for the elementary quantum of charge e] is essentially the same as the accuracy
of the most uncertain constant of radiation, k, and it surpasses by far [in accuracy] all
determinations of these quantities that have been made until now. [Planck argued, and
added:] An examination of these numbers by more direct measurements will constitute a
task for future research, which is as important as it is necessary. (Planck, 1900f, p. 245)
About eight years later Ernest Rutherford and Hans Geiger succeeded in measur-
ing the electric charge of a-particles as 9.3 x 1 0 - 1 0 electrostatic units; hence they
concluded the value of the elementary charge to be 4.65 x 1 0 - 1 0 electrostatic units,
because a-particles carry two elementary charges (Rutherford and Geiger, 1908).
'The agreement of this figure with the number calculated by me, 4.69 x 1 0 - 1 0 ,
could be taken as decisive confirmation of the usefulness of my theory,' remarked
 Max Planck and the Law of Blackbody Radiation 47

Max Planck in his Nobel lecture (Planck, 1920, p. 15). Towards the end of the first
decade of the twentieth century several different methods for determining experi-
mentally the number of molecules per gram-atom were available, as for instance the
observation of Brownian motion or critical opalescence. The results obtained from
them agreed fairly well with the theoretical value for a; - 1 (= No) which Planck had
derived in 1900 from his radiation law.
While it took several years to improve the experimental techniques to obtain
reliable determinations of the values of Avogadro's number No and the elementary
charge e, other than the ones derived by Planck from the law of blackbody radiation,
Planck himself continued with his attempts to deepen the theoretical foundations
and to elucidate the physical interpretation of his theory of blackbody radiation. In
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

a paper, which he submitted to Annalen der Physik in January 1901, Planck again
presented the theoretical derivation of his radiation formula, Eq. (18), introducing
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

certain changes in his earlier treatment (Planck, 1901a). The main difference with
the earlier derivation consisted in the fact that Planck did not obtain the magni-
tude of the discrete energy elements, e (= hi/), by comparing the entropy expression,
Eq. (19), derived from the radiation law, with the one derived from the combinato-
rial treatment, Eq. (22). He rather began with the statistical entropy arising from
the distribution of Pv energy elements of magnitude e„ among N„ resonators of
frequency v, that is,

S„ = kNv (30)

where Pvev = NVUV, and then determined the value of ev applying two well-known
and well-established laws of the theory of heat radiation: the Stefan-Boltzmann
law, Eq. (3), and Wien's displacement law, Eq. (4). Prom those laws Max Thiesen
had obtained in February 1900 the equation

Ex = T54>(\T), (31)

where ip(XT) is a function of the product of wavelength and temperature alone

(Thiesen, 1900a, p. 66). By investigating the consequence of this 'simplest formu-
lation of Wien's displacement law' (Planck, 1901a, p. 561), Planck found that the
entropy of a resonator in a cavity would depend only on the ratio of its energy and
its frequency, that is, 45
Su = f(^). (32)

Because of Eq. (32), then, the energy element in Eq. (30) had to be proportional to
the frequency, or
ev = hu. (23')

4 5 Planck had noted this consequence already when he proposed the new radiation formula,
Eq. (18), in October 1900 (Planck, 1900c, p. 201, footnote 1).
 48 The Golden Age of Theoretical Physics

And, therefore, the entropy per resonator, S„, took the form

Su 1 + hu ) (30')
\ hv ) hv hv

where h and k denote two universal constants. The expression for the (equilibrium)
energy of a resonator having frequency v, Eq. (25), and the radiation law, Eq. (18),
then followed in the same way as Planck had outlined already in December 1900
and which we have described above. 46
Planck returned to questions connected with the radiation spectrum emitted by
a blackbody, which he also called at that time the 'normal spectrum,' twice again in
1901. 47 In the first communication, presented to the Prussian Academy on 9 May,
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

he discussed the consistency of the definition of entropy, Eq. (30'), with the second
law of thermodynamics (Planck, 1901c). For this purpose he had to consider the
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

entropy of the system (cavity plus resonators) and to prove that, starting from any
distribution of the energy density, say pu, among the frequencies, the total entropy
never decreases. Planck defined the entropy of a monochromatic polarized ray of
radiation of frequency v by the expression

, , Pv \ ] n (, , P» \ P" 1. ( f>V (33)

+ 3 3 + 3 3 3 3
8nhv /c ) \ 8*hv /c ) 8nhv /c \8irhv3/c3
and studied the change of £„ during the process of absorption and reemission of
the ray by a resonator having the same frequency v. He found that the change
of the total entropy of the system (radiation plus resonator) was indeed positive
until the stationary normal density distribution was obtained, when pv assumed
the equilibrium value pu (at temperature T). 4 8
In the other paper, dedicated to the Dutch Physicist Johannes Bosscha on the
occasion of his seventieth birthday, Planck made some further comments on the
relation between entropy and probability (Planck, 1901d). In particular, he com-
pared the ways of obtaining the number of complexions in kinetic gas theory and
radiation theory. Thus he outlined first Boltzmann's derivation of the number of
possibilities, V, of distributing a given amount of energy, E, among TV molecules of
mass m, that is,
Nl
V = (34)
n[/(i>ylz>^i7>C)]«to! '
46
Planck had, of course, noticed in December 1900 that the resonator entropy, Eq. (19), was of
the form given by Eq (32).
47
T h e notion of 'normal spectrum' had occurred in Planck's publications since 1899, for the
first time in his fifth communication to the Prussian Academy on irreversible radiation processes
(Planck, 1899, Section 23). The normal spectrum was defined as the radiation spectrum obtained
in a perfectly reflecting cavity with resonators; that is, it was identical with the equilibrium
spectrum of a blackbody.
48
By means of irreversible absorption and reemission processes the polarization of the original
radiation, if any was present before, will also be removed.
 Max Planck and the Law of Blackbody Radiation 49

where f(x, y, z, £, 77, £) denotes the distribution function depending on the coordi-
nates (x, y,z) and the velocities (£,r/, £) °f * n e molecules, and da represents six-
dimensional 'elementary regions.' (Evidently, N — f fdcr and E = (m/2) x / ( £ 2 +
7 2
7 + C 2 ) ^ - ) The important point in Eq. (34) is that all elementary regions have
to have the same magnitude, which is arbitrary but chosen in such a way that
in each region a sufficiently large number of molecules is contained (Boltzmann,
1877, section II). In the case of Planck's resonators in radiation theory the number
of complexions, 1Z, could also be obtained by counting the possibilities of a cer-
tain distribution of energy; especially, one had to distribute Pv energy elements of
magnitude ev (= hv) among Nv resonators of frequency v, Pvi energy elements of
magnitude ev> (= hv) among Nv< resonators of frequency 1/, etc. The corresponding
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

number of complexions was given by the product

_{PV + NV-1)\ (P^ + i W - l )

by KAINAN UNIVERSITY on 01/10/15. For personal use only.

K }
PV\(NV - l)\ ' PV,\(NV,-1)\ ••"
If one assumes with Boltzmann and Planck that V and 1Z determine the probability
for the respective states of the systems under consideration (i.e., for the energy
distribution among the parts of the systems considered), then one may conclude:
'The entropy of a system in any state depends only on the probability of that state'
(Planck, 1901d, p. 638).
These additional remarks completed the theory of blackbody radiation, a theory
which had occupied Planck for the previous six years. He then shifted his field of
interest, partly back to problems of the theory of dissociation and solutions and
partly to other problems of radiation theory, such as the dispersion and absorp-
tion of radiation by nonconducting and conducting materials. Planck considered
the theory of heat radiation as a completed subject, and in the winter semester
of 1905-1906 he delivered lectures on it which were published as a book (Planck,
1906). Planck's Lectures on the Theory of Heat Radiation provided a systematic
and detailed presentation of all aspects of the subject. He started from the elemen-
tary optical phenomena and developed the theory by adding the electrodynamic
and thermodynamic descriptions. He devoted special attention to the statistical
definition of entropy (in Part 4) and to the demonstration of the irreversibility of
radiation phenomena (Part 5) because they seemed to throw new light on the phys-
ical interpretation of the constant h. The Lectures show that Planck had continued
to think about radiation theory since his last paper on it in 1901; he even made use
of some recent progress in the kinetic theory of matter for that purpose.
As we have emphasized above, the central role in establishing the blackbody
radiation formula, Eq. (18), was played by the relation between entropy and proba-
bility of the stationary state of the system consisting of radiation and resonators. In
1900 Planck had based his approach completely on the procedure which Boltzmann
had developed in 1877. Then, two years later, he received a new book entitled Ele-
mentary Principles in Statistical Mechanics; it was sent to him by the author, Josiah
Willard Gibbs, himself, then professor at Yale University, New Haven, Connecticut
 50 The Golden Age of Theoretical Physics

(Gibbs, 1902). 49 Planck esteemed Gibbs highly because of his important contri-
butions to the thermodynamic description of phase transitions; hence he was very
interested to read what Gibbs had to say about kinetic theory, especially about
those parts which constituted the foundation of his radiation theory. 50 Gibbs had
based his treatise, to which he gave the qualifying subtitle 'with especial reference to
the rational foundation of thermodynamics,' on considerations of the phase space;
that is, he had studied the description of dynamical systems of n degrees of free-
dom in a 2n-dimensional space of position and momentum variables, q\,..., qn and
pi,...,pn. All thermodynamic properties of the system could be obtained from the
distribution of the variables in phase space. Thus Gibbs arrived, for example, at
a definition of the entropy of a system having the kinetic energy Ekin as (Gibbs,
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

1902, Chapter XIV),

S = klnV, (36)
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

where V represents the volume of phase space,

V = J---Jdq1...dpn, (37)

occupied by all systems of the same or smaller kinetic energy (i.e., p\/2m -\-p\j2ra
-\ 1- p\l1m < Ekin, if m is the mass of the particles in the system).
Planck, though he admired the generality and elegance of Gibbs' presentation,
found that for practical purposes his definition of entropy was not as useful as it
pretended to be. In a paper on the mechanical interpretation of temperature and en-
tropy, in which he compared Boltzmann's definition of entropy via the complexions
(1877) and Gibbs' definitions (1902), Planck complained that Gibbs — in contrast
49
Josiah Willard Gibbs was born on 11 February 1839 in New Haven, Connecticut. Beginning
in 1854 he studied at Yale University, graduating in 1858. Upon receiving his doctorate at Yale
in 1863 he was appointed a tutor for three years. He continued his studies abroad: Paris (1866-
1867), Berlin (1867-1868), and Heidelberg (1868-1869); at Heidelberg he studied with Kirchhoff
and Helmholtz. He returned to New Haven in 1869 and two years later was appointed Professor of
Mathematical Physics at Yale University, a position which he occupied until he died on 28 April
1903 in New Haven.
Gibbs wrote his first paper on the thermodynamics of fluids in 1873. He continued to work on
thermodynamical problems, contributing in 1876 and 1878 three memoirs on the equilibrium of
heterogeneous substances, which made him famous among the specialists in America and Europe.
Besides thermodynamics, he worked on electromagnetic theory of light and developed vector anal-
ysis. The book on kinetic theory, Elementary Principles in Statistical Mechanics (Gibbs, 1902),
represented his contribution to the bicentennial celebration of Yale College.
Gibbs was a member of many learned societies including the U.S. National Academy of Sciences,
the Royal Society of London, the Royal Institution of Great Britian, the Prussian Academy of Sci-
ences and the Gesellschaft der Wissenschaften zu Gottingen. He received honorary degrees from
Erlangen, Christiana and Princeton Universities and was awarded the Rumford Medal (1881) of
the American Academy of Arts and Sciences, Boston, and the Copley Medal (1901) of the Royal
Society.
Gibbs sent copies of his book on Statistical Mechanics to many of the well-known physicists
in the United States and Europe, including Lord Rayleigh, H.A. Lorentz and Max Planck, who
acknowledged receipts of the book. It was translated into German by Planck's former assistant
Ernst Zermelo and published in 1905.
 Max Planck and the Law of Blackbody Radiation 51

to Boltzmann — had not given any prescription for caculating t h e entropy of a

nonequilibrium s t a t e .
Therefore I believe [I am] allowed to state the result of this investigation as follows [he
concluded in July 1903:] For the overwhelming generality with which, as it strikes one on
first inspection, Gibbs has formulated his various definitions of entropy — i.e., without
referring to the nature of the systems considered — he has to pay by restricting their
physical significance. For all reversible processes, Gibbs' definitions will serve as well as
several other already existing definitions of a more formal nature. However, for irreversible
processes, which alone endow entropy with its full significance and which provide the key to
a complete understanding of thermal equilibrium, Boltzmann's definition of the entropy
turns out to be the most appropriate and productive among all those known hitherto.
(Planck, 1904b, pp. 121-122)
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

A b o u t a year later Planck discovered t h a t Gibbs' entropy definition, Eq. (36), was
very useful in radiation theory, for it provided a physical interpretation of the con-
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

s t a n t h occurring in his radiation law. He immediately presented this interpretation

in t h e lectures on heat radiation (Planck, 1906, Section 150, p p . 154-156).
T h e m e t h o d of the phase space allowed one, as Planck recognized, to derive
the entropy of t h e resonator directly from its dynamical variables in t h e following
way. T h e probability t h a t the energy of t h e resonator assumes values between E
and AE is given according to Eq. (37) 'by t h e m a g n i t u d e of t h a t area of t h e s t a t e
variables / a n d g, which is bounded by t h e curves E — const, a n d E + AE = const.'
(Planck, 1906, p . 155). T h u s Planck represented the resonator by a system of two
charges of opposite sign which oscillate towards each other along a fixed line. T h e
Hamiltonian H of t h e system is given by

H = E=l-Kf2+l-Lf\ (38)

where / is the product of t h e positive charge times t h e distance from t h e centre of

the dipole a n d / its time derivative; K a n d L denote t h e constants of t h e system.
Now / a n d g = Lf (= dH/df) constitute a pair of canonically conjugate dynamical
variables of the linear oscillator or resonator. According to Gibbs t h e s t a t e of the
latter may be determined by points in t h e two-dimensional (/, <7)-phase space. T h e
curves E = const, and E = const. + AE are ellipses having t h e area E/v a n d
(E + AE)/v, respectively; hence the difference of t h e two areas in phase space
becomes AE/i/.51 Planck now noted:

If we assume that the entire phase-plane ['Zustandsebene,' i.e., the phase space of the
one-dimensional resonator] is divided by a large number of these ellipses into seperate
sections, such that the ring-shaped areas bounded by two successive ellipses become equal
in magnitude to one another, such that AE/u = const., then we obtain a method of
determining those sections AE of the energy, which correspond to equal probability and
which we may therefore call energy elements. [And he continued:] If we put the magnitude

51
T h e semiaxes of the ellipse, E = const., are y/2E/K and \p2EL\ the area becomes 2KE^/L/K
or E/v, since the frequency is given by v = (l/27r) \JK/L.
 52 The Golden Age of Theoretical Physics

of an energy element AE equal to e, and put the constant in the previous equation [AE/u =
const.] equal to h, then we arrive exactly at the former Eq. [(23')] without having made use
of Wien's displacement law. At the same time the elementary quantum of action h acquires
a new meaning, namely it gives the area of an elementary region in the phase-plane of a
resonator, no matter what its frequency. (Planck, 1906, pp. 155-156)

In other words, Gibbs' entropy definition clarified the concept of t h e energy elements
in radiation theory by identifying the constant h in Eq. (23') with t h e elementary
region in t h e two-dimensional phase space of the resonator. T h e same h a p p e a r s
in t h e equation for the energy elements of arbitary frequency v. Morever, if the
resonator is replaced by any other electrodynamic system (of one degree of freedom),
which is capable of absorbing and emitting radiation such t h a t a n equilibrium can
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

be established with radiation in a cavity, t h e n the s a m e constant h determines

the elementary regions in its phase space. Hence Planck took this constant as
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

representing a universal feature of n a t u r e a n d rightly n a m e d h as the 'elementary

q u a n t u m of action.' 5 2
While t h e relation between entropy and t h e probability of the resonator's s t a t e
together with t h e hypothesis of t h e elementary q u a n t u m of action provided t h e
correct description of the stationary (i.e., equilibrium) s t a t e in a cavity, it did not
yet guarantee t h a t the final s t a t e was actually approached. To some extent t h e
hypothesis of n a t u r a l radiation — i.e., the assumption t h a t radiation in the original
s t a t e consists of a disordered mixture of plane waves having arbitary phases a n d
directions (which form a cone of finite a p e r t u r e ) — m a d e sure t h a t by t h e interaction
of this radiation with resonators t h e entropy is actually increased; especially, the
intensity a n d polarization of the radiation of a given frequency will be changed t o
complete uniformity. 5 3 Still, the available resonator mechanisms of absorption and
emission of radiation (having a given frequency) could not establish a n exchange
of energy w i t h radiation of different frequencies, a process which is necessary in
order to o b t a i n t h e stable final equilibrium distribution. Planck suggested t h a t the
collisions of resonators might imply frequency-changing interactions, a n d called for
a b e t t e r theoretical description in the future.

From such a theory [he concluded], one must then surely expect to obtain far-reaching
information about the constitution of oscillators that exist in nature: this is motivated by
the fact that the theory must, in any case, also yield a deeper explanation of the physical
significance of the universal quantum of action h, an explanation of the same importance
as that of the elementary quantum of electricity. (Planck, 1906, pp. 127-128)

Planck's Lectures on the Theory of Heat Radiation represented t h e rational for-

mulation of his vision of a unified theory of all radiation phenomena. T h e basis
52
Planck introduced the notion of the 'elementary quantum of action' with the following words:
'I want to designate this [i.e., the constant h] as the "elementary quantum of action" ["e/e-
mentares Wirkungsquantum"] or as "element of action" ["Wirkungselemenff'], because it has the
same dimension as the qunatity which owes its name to the Priniciple of Least Action' (Planck,
1906, p. 154).
53
One cannot speak of the entropy of a plane electromagnetic wave.
 Max Planck and the Law of Blackbody Radiation 53

for this unification was provided by thermodynamics, which accounts for b o t h t h e

reversible a n d irreversible process in n a t u r e . In radiation of theory t h e p h e n o m e n a
of propagation, reflection a n d diffraction of light could b e described as reversible
processes, while the phenomena of absorption and emission of light a n d the transfor-
mation of frequencies seemed t o require a description as irreversible processes. As
soon as Planck h a d written down his unified conception of radiation, it was attacked
severely from two sides: O n t h e one hand, Albert Einstein destroyed the unity of
heat radiation by emphasizing the qualitatively different n a t u r e of short and long
wavelength radiations; on the other, closer investigations showed t h a t t h e radiation
formula, Eq. (18), contradicted the fundamental concepts of existing electrodynamic
theory. I shall discuss these developments in other essays.
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

References
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

A. Bartoli, 1876, Sopra il movimenti dalla luce e dal calore, Florence: Le/Mounier.
L. Boltzmann, 1872, Sitz. ber. Akad. Wiss. (Wien) 66, 275-370.
L. Boltzmann, 1877, Sitz. ber. Akad. Wiss. (Wien) 76, 373-435.
L. Boltzmann, 1884a, Ann. d. Phys. (3) 22, 31-39.
L. Boltzmann, 1884b, Ann. d. Phys. (3) 22, 291-294.
L. Boltzmann, 1895b, Nature 5 1 , 413-415.
L. Boltzmann, 1895c, Vorlesungen iiber Gastheorie. I, Leipzig: J.A. Barth.
L. Boltzmann, 1897 Sitz. ber. Preuss. Akad. Wiss. (Berlin), 600-662.
C. Christiansen, 1884, Ann. d. Phys. (3) 2 1 , 364-369.
W. Friedrich, P. Knipping and M. von Laue, 1912, Sitz. ber. Bayer. Akad. Wiss.
(Miinchen), 303-322.
J.W. Gibbs, 1902, Elementary Principles in Statistical Mechanics, New Haven: Yale Uni-
versity Press.
H. Hertz, 1888a, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 197-210; Ann. d. Phys. (3)
34, 551-569.
H. Hertz, 1888b, Ann. d. Phys. (3) 36, 1-22.
G.G. Hettner, 1922, Naturwiss. 10, 1033-1038.
H. Kangro, 1970, Vorgeschichte des Planckschen Wirkungsquantums, Wiesbaden: Steiner
Verlag. English translation: Early History of Planck's Radiation Law, London:
Taylor & Francis, 1976.
G. Kirchhoff, 1859a, Monatsber. Preuss. Akad. Wiss. (Berlin), 662-665; Gesammelte
Abhandlungen, 567-570.
G. Kirchhoff, 1859b, Monatsber. Preuss. Akad. Wiss. (Berlin), 783-787; Gesammelte
Abhandlungen, 567-570.
G. Kirchhoff, 1860, Ann. d. Phys. (2) 109, 275-301; Gesammelte Abhandlungen, 571-598.
G. Kirchhoff and R. Bunsen, 1860, Ann. d. Phys. (2) 110, 160-189; Phil. Mag. (4) 20,
89-109.
T. Kuhn, 1978, Black-Body Theory and the Quantum Discontinuity, 1894-1912, The
Clarendon Press, Oxford.
S.P. Langley, 1881, The Chemical News and Journal of Physical Science 4 3 , 6.
S.P. Langley, 1886, Phil. Mag. (5) 21, 394-409.
J. Leslie, 1804, An Experimental Inquiry into the Nature and Propagation of Heat, London:
J. Mawman.
E. Lommel, 1878, Ann. d. Phys. (3) 3, 251-283.
 54 The Golden Age of Theoretical Physics

H.A. Lorentz, 1895, Versuch einer Theorie der elektrischen and optischen Erscheinungen
in bewegten Korpern, Leyden; E.J. Brill.
O. Lummer and E. Jahnke, 1900, Ann. d. Phys. (4) 3, 283-297.
O. Lummer and F. Kurlbaum, 1898, Verh. d. Phys. Ges. Berlin 17, 106-111.
O. Lummer and E. Pringsheim, 1899a, Verh. d. Deutsch. Phys. Ges. (2) 1, 23-41.
O. Lummer and E. Pringsheim, 1899b, Verh. d. Deutsch. Phys. Ges. (2) 2, 23-41.
O. Lummer and E. Pringsheim, 1900, Verh. d. Deutsch. Phys. Ges. (2) 1, 23-41
(presented at the Naturforscherversammlung, Aachen, 18 November 1899).
O.E. Meyer, 1899, Die Kinetische Theorieder Gase, second edition, Breslau; Maruschke
und Berendt.
W. Michelson, 1888, Phil. Mag. (5) 25, 425-435.
F. Paschen, 1896, Ann. d. Phys. (3) 58, 455-492.
F. Paschen, 1899a, Sitz. her. Preuss. Akad. Wiss. (Berlin), 405-420.
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

F. Paschen, 1899b, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 959-976.

F. Paschen, 1901, Ann. d. Phys. (4) 4, 277-298.
M. Planck, 1879, Physikalische Abhandlungen and Vortrage 1, 1-61.
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

M. Planck 1881, Ann. d. Phys. (3) 13, 535-543; Physikalische Abhandlungen and Vortage
1, 125-133.
M. Planck, 1890, Ann. d. Phys. (3) 40, 561-576.
M. Planck, 1894a, Verh. d. Phys. Ges. Berlin 13, 9-29.
M. Planck, 1894b, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 661-644.
M. Planck, 1894c, Sitz. ber. Bayer. Akad. Wiss. (Munchen) 24, 391-394.
M. Planck, 1895a, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 289-301.
M. Planck, 1895b, Ann. d. Phys. (3) 57, 72-78.
M. Planck, 1896, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 151-170.
M. Planck, 1897a, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 57-68.
M. Planck, 1897b, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 715-717.
M. Planck, 1897c, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 1122-1145.
M. Planck, 1898, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 449-476.
M. Planck, 1899, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 440-480.
M. Planck, 1900a, Ann. d. Phys. (4) 1, 69-122.
M. Planck, 1900b, Ann. d. Phys. (4) 1, 719-737.
M. Planck, 1900d, Verh. d. Phys. Ges. (2) 2, 202-204.
M. Planck, 1900e, Ann. d. Phys. (4) 3, 746-766.
M. Planck, 1900f, Verh. d. Deutsch. Phys. Ges. (2) 2, 237-245.
M. Planck, 1901a, Ann. d. Phys. (4) 4, 553-563.
M. Planck, 1901c, Sitz. ber. d. Preuss. Akad. Wiss. (Berlin), 544-555; Ann. d. Phys.
(4) 6, 818-831.
M. Planck, 1901d, Ann. d. Phys. (4) 9, 619-628 (1902); Jubilee Volume for J. Bosscha,
Archive Nerlandaises, 55-66.
M. Planck, 1906, Vorlesungen iiber die Theorie der Warmestrahlung, Leipzig: J.A. Barth.
M. Planck, 1920 (Nobel lecture), Leipzig: J.A. Barth.
M. Planck, 1923a, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 66-74.
M. Planck, 1943, Naturwiss. 3 1 , 153-159.
M. Planck, 1948a, Wissenschaftliche Selfstbiographie, Leipzig: J.A. Barth.
Lord Rayleigh, 1900b, Phil. Mag. (5) 49, 539-540.
F. Richarz, 1894, Sitz. ber. Bayer. Akad. Wiss. (Munchen); Ann. d. Phys. (3) 52,
385-416.
H. Rubens and E.A. Aschkinass, 1899, Ann. d. Phys. (3) 65, 241-256.
H. Rubens and F. Kuilbaum, 1900, Sitz. ber. Preuss. Akad. Wiss. (Berlin), 929-941.
 Max Planck and the Law of Blackbody Radiation 55

H. Rubens and E.F. Nichols, 1896, Naturwissenchaftliche Rundschau 11, 545-549.

E. Rutherford and H. Geiger, 1908, Proc. Roy. Soc. (London) A 8 1 , 162-173.
K. Scheel, 1913, Naturwiss. 1, 177-180.
B. Schwalbe, 1900, Verh. d. Deutsch. Phys. Ges. (2) 2, 147-159.
D.M. Siegel, 1976, Isis 67, 565-600.
A. Sommerfeld, 1919, Atombau und Spektrallumien, Braunschweig: Pr. Vieweg & Sohn.
J. Stefan, 1879, Sitz. ber. Akad. Wiss. (Wien) 79, 391-428.
B. Stewart, 1858, Transactions of the Royal Society of Edinburgh 22, 1-20.
M. Thiessen, 1900a, Verh. d. Deutsch. Phys. Ges. (2) 2, 65-70.
J.J. Thomson, 1898b, Phil. Mag. (5) 46, 528-545.
W. Wien, 1892, Ann. d. Phys. (3) 45, 685-728.
W. Wien, 1893a, Site. ber. Preuss. Akad. Wiss. (Berlin), 55-62.
The Golden Age of Theoretical Physics Downloaded from www.worldscientific.com

W. Wien, 1893b, Ann. d. Phys. (3) 49, 633-641.

W. Wien, 1894, Ann. d. Phys. (3) 52, 132-165.
W. Wien, 1896, Ann. d. Phys. (3) 58, 662-669.
by KAINAN UNIVERSITY on 01/10/15. For personal use only.

W. Wien, 1900a, Rapports presentes au Congres International de Physique reuni a Paris

en 1900, Paris: Gauthier-Villars, Vol. 2, 23-40.
W. Wien, 1900b, Ann. d. Phys. (4) 3, 530-539.
W. Wien, 1930, Wichelm Wien: Aus dem Leben und Werken eines Physikers (K. Wien,
ed.), Leipzig: J.A. Barth.
W. Wien and O. Lummer, 1895, Ann. d. Phys. (3) 56, 451-456.
E. Zermelo, 1896, Ann. d. Phys. (3) 57, 485-494.