A short history of Polish mathematics

by W. Żelazko (Warszawa)

In this article I shall explain how Poland - a country having at the beginning of 20th
century almost no mathematical traditions - could achieve, within a relatively short period
1919-1939, a good international position in such fields of mathematics as functional analy-
sis, topology, set theory, functions of a real variable, logic and foundations of mathematics,
and good forecasts for the future development of probability theory, differential equations
and Fourier analysis.
The reason why we had poor mathematical traditions is rather clear. In 19th century
– a period of a great development of mathematics in Western Europe – Poland was not
an independent country. In 1795 it was partitioned between Austria, Russia and Prussia
(Germany was not yet unified), and the independence was retrieved only at the end of
1918. Thus in 19th century the essential effort of the nation was set onto humanities, since
literature and poetry were necessary for supporting the idea of independence and even for
preservation of the language (in some periods the children in Prussian or Russian parts
were taught only in German or Russian and Polish language was forbidden; the situation
in the Austrian part was much more liberal). For that reason the mathematical papers
were written only in Polish, so even if there were some original results, they remained not
recognized by the mathematical community. Before the first World War (1914-18) there
were in the Russian part of Polish territories two universities, a Russian University in
Warsaw, offering lectures only in Russian (evacuated at the outburst of the war to Rostov-
on Don, where it still exists; the previous University of Warsaw, with Polish professors,
was closed after the upraising against Russians in 1863). Mathematics was taught there
by a quite good Russian mathematician G. Voronoy, working mostly in number theory.
The other university in Wilno (now Vilnius in Lithuania), was of much lower level. In
the Prussian part of Poland, there was no university at all. Much better situation was
in the Austrian part with two universities in Kraków (Cracow) and in Lwów (now Lviv
in Ukraine). In the University of Kraków (founded in 1364 - the oldest in Poland, older
than any German or Austrian University) there were at the beginning of 20th century
two quite good Polish mathematicians: Kazimierz Żorawski (1866-1953), with Ph.D. made
in Germany in Leipzig under Sophus Lie, and working in the differential geometry, and
Stanislaw Zaremba (1863-1942), who obtained his Ph.D. in Paris. Zaremba specialized in
partial differential equations, potential theory and mathematical physics. Also in Lwów
Univeristy there were two good mathematicians: Józef Puzyna (1856-1919) who specialized
in complex analysis and integral equations, and Waclaw Sierpiński (1882-1969). Sierpiński
finished the Russian University in Warsaw in 1904 (under G. Voronoy), made his Ph.D.
in 1906 in the University of Kraków and moved to Lwów, where he got a position at the
University, first as a lecturer and since 1910 as a professor. Sierpiński at the beginning
of his mathematical career worked in the classical number theory and later moved to the
set theory (in 1909 he gave a university course of this theory, it seems to be first such a
course ever given). After the outbreak of the first World War Russian troops took Lwów

1
and Sierpiński was internated first in Vyatka and later in Moscow, where he met several
Russian mathematicians such as Egorov or Lusin. This meeting led to future collaboration
of the Polish and Russian Schools of Mathematics. In turn, about the same time Warsaw
was taken by the Prussian troops, and Prussians agreed to reopen in 1915 the Warsaw
University. The chairs of mathematics were taken by two young mathematicians, Stefan
Mazurkiewicz (1888-1945) and Zygmunt Janiszewski (1988-1920), both were topologists.
Mazurkiewicz obtained his Ph.D. in Lwów under Sierpiński and Janiszewski in Paris in
1911. The board which passed his thesis consisted of H. Poincaré, H. Lebesgue and M.
Fréchet. An important rôle in creation of the Polish Mathematical School was played by
Zygmunt Janiszewski. Still before the end of the war it was clear that Poland will obtain
independence, people were very entusiastic about it and made lot of plans for the future.
Such a plan for mathematics was proposed by Janiszewski in his article ”On the needs
of mathematics in Poland”. His main claim was that we have no chances in the well es-
tablished theories, since we had no traditions and no great knowledge. However in the
new fields,just emerging, we have the same chances as everybody else. Thus we should
concentrate on the modern fields of mathematics (in this time it was set theory, topol-
ogy, theory of functions of a real variable and logic with foundations of mathematics),
and establish an international journal devoted only to these fields. The first volume of
this journal, Fundamenta Mathematicae, appeared in 1920 and it was the first specialized
mathematical journal in the world. Janiszewski could see this journal only in the galley-
proofs, he died from an epidemic disease earlier in the same year at the age of only 32 years.
After Poland obtained independence in November 1918, the Warsaw University had two
more professors connected with mathematics: Jan Lukasieicz - a philosopher and logician
(among his students were world famous mathematicians working in foundations of math-
ematics: Alfred Lindenbaum, Andrzej Mostowski and Alfred Tarski), and Sierpiński, who
after internation in Russia did not return to Lwów, but to Warsaw – his home city. Many
young talented students entered Warsaw University, the famous in the future mathemati-
cians such as, Bronislaw Knaster, Kazimierz Kuratowski, Szolem Mandelbrojt, Stanislaw
Saks, Alfred Tarski, Tadeusz Ważewski, Antoni Zygmund, and later Karol Borsuk, Samuel
Eilenberg, Edward Szpilrajn (Marczewski), and many others. The Warsaw topologists
(Sierpiński, Mazurkiewicz, Kuratowski, Knaster, later Borsuk and Eilenberg) had many
important results in the theory of continua and curves, compacts and fix-point theory
and in general topology (the system of axioms based upon the closure operation proposed
by Kuratowski in his Ph.D. thesis and developed in [7] is now in common use). Widely
known are Sierpiński’s ”carpet” and Borsuk’s theory of retracts (applied later by Ważewski
to differential equations). The term ”Polish spaces”, given by French mathematicians to
complete separable metric spaces also illustrates the work of Polish topologists.
Important results were obtained also in the set theory. Sierpiński, Kuratowski, Tarski
and others worked on basic questions of this theory, such as properties of ordinal numbers,
the continuum hypothesis (see e.g [11]) and the axiom of choice. The Kuratowski’s ”max-
imum principle”, rediscovered 13 years later by Max Zorn, is called now Zorn lemma or
the Kuratowski-Zorm lemma.
The efforts of mathematicians working in real functions culminated later in famous
books of Saks [10] and Zygmund [14].

2
 The important feature of the Warsaw School was close cooperation of its participants.
They discussed topics between themselves, presented them on the meetings of the Polish
Mathematical Society and often published joint papers. Such a situation attracted to
mathematics many young people, who also tried to obtain and publish new results even
during the university studies. It was in contrast with the situation in Kraków, where
Zaremba and Żorawski worked separately and no mathematical school was formed before
the World War II.
The other, perhaps even more important mathematical school, was created in Lwów,
due to Banach, Steinhaus with their students and collaborators. The story of Stefan
Banach is quite extraordinary (for more details see Steihaus [12] and Kaluża [6] some
details in these publications do not coincide). In 1916, when sitting in a Kraków park, Hugo
Steinhaus overheard a conversation between two young people talking about the Lebesgue
integral. Since the concept of this integral was quite fresh, Steinhaus approached to them
and started a conversation. The young men were Stefan Banach, a student of the Lwów
Polytechnic Institute (the studies were interrupted by the war) and Otto Nikodym (of
the Radon-Nikodym theorem). In their talk Steinhaus mentioned the problem in Fourier
series, he was working on in this time. It was already known that for a function f in
L2 (0, 2π) the partial sums sn (f ) of its Fourier series are tending to f in the L2 -norm,
and Steinhaus tried to work out whether a similar result is true for the space L1 (0, 2π),
∫ 2π
i.e. whether for f ∈ L1 (0, 2π) we have limn 0 [f (t) − sn (f )(t)]dt = 0. To his surprise,
a few days later Banach constructed a counterexample, a function f in L1 (0, 2π) such
that limi ||f − sni || → ∞ for a certain sequence (ni ) of integers. Their joint paper was
submitted to the Bulletin of Kraków Academy of Sciences and Letters and published, due
to war, only in 1919. It was the first paper of Banach. Due to the war and lack of scientific
information the authors did not know that the problem was already solved by Hahn. Since
this time Steinhaus often discussed mathematics with Banach, Nikodym and their friend
Wilkosz, who later became a professor of mathematics in the University of Kraków.
Stefan Banach was born on 30 March 1892. According to Steinhaus [12], a railroad
employee, named Greczek brought a small baby to a Kraków laundress named Banach,
gave her a certain amount of money and asked her to keep the baby. Greczek was father of
Banach and the name of his mother is not known (perhaps it was a lady from aristocracy;
my friend Marceli Stark told me once that he knew who was the Banach’s mother, but that
he promised never to tell it anybody), thus we should be rather speaking about Greczek
spaces ! It was established later that the name of the laundress was different, and it seems
that Banach was the name of a servant girl who agreed to testify that Banach was her son
(in [6] it is stated that this girl was a real mother of Banach, what follows from suitable
documents). When Banach was about 15 years old he had to care for himself getting money
by tutoring school pupils in mathematics. In 1910 he finished the secondary school and
(probably, the date is not sure) entered the Lwów Polytechnic Institute. At the outbreak
of war in 1914 the studies were interrupted and Banach returned to Kraków. He has passed
all necessary examinations for first two year of studies and it was the end of his formal
education.
Banach dreamed about an assistentship in mathematics in his school – the Polytechnic
Institute in Lwów, and thanks to Steinhaus it could be arranged. In 1920 Banach become

3
an assistant at the chair of mathematics occupied by professor Lomnicki. The same year
he obtained his Ph.D. in spite of having no formal university studies. It was possible by a
special permission from the Ministry of Education. His dissertation, translated to French,
was published in Fundamenta Mathematicae ([1]). In this paper Banach gave axioms
for the Banach spaces (the term itself was introduced later by M. Frechét and was not
in use for a long time). In an article in the first volume of Studia Mathematica (1929)
Steinhaus writes ”Wir nennen – mit Herr Fréchet – einen Banachschen Raume, kurz einen
B-Raum...”. In his book [3], or its French version [4] Banach calls these spaces ”spaces of
type (B)”, and the term ” Banach space” is used directly only in 1940 in the mentioned
below paper of M. Eidelheit. In his thesis Banach established basic properties of bounded
linear operators between such spaces. There is also the Banach fix point theorem with
an application to integral equations. Two years later Banach obtained his habilitation
(published later in [3]) and in the same year obtained the position of an extraordinary
professor in the University of Lwów. In his habilitation thesis Banach has solved a problem
of Lebesgue. The question was whether one can construct on Rn a (positive) ”measure”
which is equal to one on the unit cube, is invariant under the isometric maps and is defined
on all subsets of Rn . Of course, such a ”measure”, if existed, could not be countably
additive, so that only the finite additivity was postulated. By an example of Hausdorff it
was known that even such a measure cannot exist for n ≥ 3, but the problem was open
for n = 1 or 2. And Banach solved this problem in positive. Such a measure does exist on
the line or plane. In the construction Banach used a developed later (see [4]) concept of
generalized limit (based upon a preliminary version of the Banach-Hahn theorem). From
this work emerged the concept of a Banach mean given later by John von Neumann. This
work was completed by an interesting result obtained in a joint paper with Tarski [5].
Call two subsets A and B of Rn equivalent by means of a finite decomposition,
∪k ∪ifk we can
decompose both sets into a finite number of disjoint parts A = 1 Ai , B = 1 Bi and
the parts Ai and Bi are mutually isometric. Their result is as follows: Suppose that A
and B are two bounded subsets of Rn , n ≥ 3 with non-void interiors. Then these sets are
equivalent by means of a finite decomposition. Usually this result is formulated for the case
when A is a ball and B is the union of two disjoint balls of the same radius, but the result
is much more general. In the proof the authors make use of a construction given earlier
by Hausdorff, which showes, in particular, that for n ≥ 3 the mentioned above problem of
Lebesgue has a negative solution.

Banach quite soon attracted the attention of many good students and young mathe-
maticians. They used to meet in a cafeteria and during long sessions discussed mathemat-
ics. In these meetings participated also the professors and sometimes the foreign guests.
Soon formed a group of scientists working not only in functional analysis, the famous Lwów
school of mathematics. It was closely connected with the Warsaw school, especially after
1927, when Kuratowski obtained a chair in Lwów Polytechnic Institute. Let me mention
several names connected with this school. Stanislaw Mazur, 1905-1981, was one of main
collaborators of Banach. In particular he helped Banach in preparation the book [4], or
its Polish version [3]. Mazur himself is known for numerous results in functional analysis
(one of his classical results says that the convex envelope of compact subset of a Banach
space is again compact) and summability theory, and widely known is the Mazur-Gelfand

4
theorem. Mazur also collaborated closely with Wladyslaw Orlicz (1903-1990) obtaing re-
sults in non-linear functional analysis (polynomial operators) and in locally convex spaces
(Bo -spaces, but their results on these spaces were published only after World War II).
Orlicz is known for his results on orthogonal series, real functions, summability theory
and functional analysis (polynomial operators, interpolation of operators and theory of
modular spaces). After his name comes the concept of Orlicz Spaces. Very important
mathematician in this group was Julian Schauder, 1899-1943. With his name there are
connected such concepts and results as a Schauder basis, Schauder fix-point theorem and
Lerey-Schauder theory of non-linear completely continuous operators (with topological
methods based upon the concept of an index and with applications to partial differential
equations). Herman Auerbach, (perished in 1942), is known for his results in the theory of
convex bodies. He had an encyclopedic memory and knew lot of mathematical literature.
So whenever Banach gave a subject to a student he was sending him to Auerbach for suit-
able references. Max (Meier) Eidelheit had only a few results before his premature death
in 1943. Let me mention one of them, published in Studia Math. in 1940. Let X and Y be
two Banach spaces and suppose that the algebras L(X) and L(Y ) of bounded operators
are algebraically isomorphic. Then the spaces X and Y are topologically isomorphic. It
is one of first results in so called automatic continuity. One should mention also Stefan
Kaczmarz (perished in 1939), working in the theory of orthogonal series (he published with
Stenhaus a book in this subject), and much younger Stanislaw Ulam, 1909-1984, the first
Ph.D. of Kuratowski, and Marek (Mark) Kac (1914-1984) - with his Ph.D. made under
Steinhaus. Both of them emigrated before 1939 to the United States and played there an
important rôle (Ulam participated in the Los Alamos team working in construction of the
nuclear weapon).
In 1928 Banach and Steinhaus founded in Lwów a new international journal ”Studia
Mathematica”. The first volume appeared in 1929 and contained 14 papers. Four of them
(by Kaczmarz, Steinhaus and two papers of Orlicz) were devoted to orthogonal series –
which shows the current interest of the Lwów School in late tweenties. This volume contains
also the Hahn-Banach theorem (in the second volume of Studia Math. Banach stated that
a similar result was published by Hahn in 1927), the Banach’s theorem on the continuity
of an inverse operator, and the result of Mazur stating that all spaces Lp (0, 1), p ≥ 1, are
mutually homeomorhic.
In 1931 the famous series ”Monografie Matematyczne” (Mathematical Monographs,
shortly M.M.) started with the classical book of Banach [4]. Many results, such as the
closed graph theorem, the continuity of inverse of a linear operator, and the Banach-
Steinhaus theorem were obtained in this book in a more general setting of an F -space,
i.e. a completely metrizable topological vector space (Banach used this letter to honour
M. Fréchet, and in this sense this term is used by the Polish School; later French math-
ematicians added the condition of local convexity and call such spaces Fréchet spaces).
Other books in this series are positions [7], [10] and [11] on the reference list, as well as
the mentioned above book of Kaczmarz and Steinhaus.
As it was mentioned earlier, the Lwów mathematicians almost daily gathered in a
cafeteria, discussed mathematics and proved theorems. The cafeteria was called Kawiarnia
Szkocka (Scottish Café). The sessions could last there for several hours ending late in the

5
night or even early morning (for a detailed description see [13] and [9]). During one
night they have obtained a particularly interesting result. However, in the morning all
their notes were wipped out from a marble table (they used pencils), and they could not
reconstruct neither the proof nor even the result itself. Consequently Mrs Banach bought
a copybook, gave it to a waiter and asked to hand it to the mathematicians whenever
they appeared. In this way arose the famous Scottish Book. The first entry in this book
is done by Stefan Banach on 17 July 1935 and last, No. 193, by Hugo Steinhaus on 31
May 1941. The mathematicians used to write there problems and sometime solutions.
Usually for solving problems there were offered some prizes. For instance Mazur offered
a live goose for solution of an approximation problem (known to be equivalent with the
problem of existence of a Schauder basis in every Banach space). The goose was won by
Per Enflo only in 1972 for the negative solution of this problem. The cafeteria was visited
also by foreign guests (some entries in the Scottish Book are given by M. Fréchet, J. von
Neumann, C. Offord and others).

The future plans of the Polish School were quite ambitious. It follows from the title
of the Polish version of the Banach’s book (see [3]), as well as from some papers of Mazur
and Orlicz, and of Lerey and Schauder, that they planned to develop a theory of non-linear
operators within functional analysis. The Polish Mathematical Society planned to organize
two mathematical institutes, one in Lwów – devoted to applied mathematics, and one in
Warsaw – devoted to the pure mathematics. Polish mathematicians realized that their
mathematics is quite one-sided and rather narrow. In particular some classical theories, as
the theory of analytic functions or algebra almost did not exist in Poland. They wanted
to develop these theories too and also call more attention to applications of mathematics.
There were some possibilities for developping more classical analysis. In 1930 Antoni Zyg-
mund obtained a chair of mathematics in the University of Wilno, and soon got a very
gifted student Józef Marcinkiewicz (1910-1940). Marcinkiewicz finished university in 1933,
got his Ph.D. under Zygmund in 1935 and habilitated in 1937. In 1935/36 he stayed in
Lwów working with Kaczmarz and Schauder and in 1939 was invited for a chair of math-
ematics to the Poznań University (founded in 1919), where, since 1937, Wladyslaw Orlicz
already was a professor. In 1939 Marcinkiewicz had already many published papers in
real functions (singular integrals, differentiability of integrals, interpolation polynomials),
trigonometric series, orthogonal systems, complex analysis, probability theory and theory
of operators (interpolation theorems). At the outbreak of the World War II Marcinkiewicz
was in Great Britain, working with Lttlewood, but immediately came back to Poland,
served in the army, was taken prizoner of war by Sovjets (on the 1st of September 1939
Poland was invaded by Germany and, according to the Ribbentrop-Molotov agreement, on
17 of September by Sovjet Union, so that soon lost its independence and was partitioned
again between Germany and Sovjet Union), and executed (probably in Katyń) together
with several thousands of other Polish officers and civil servants. Antoni Zygmund writes
that the death of Marcinkiewicz was the heaviest individual lost of Polish Mathematics
during the second world was, he told me once that if Marcinkiewicz could live longer he
would be one of greatest mathematicians of 20th century, perhaps better than Stefan Ba-
nach. In spite of a young age the list of his publications is about the same size as the list of
Banach’s publications, also his influence on future work in mathematical analysis is quite

6
substantial.
In the German part of Poland (it included Warsaw, Kraków and Poznań) not only
Universities, but also secondary schools were closed. Poland was supposed to provide
Germany with an unskilled cheap labour of uneducated people. Since Germany attacked
first, many mathematicians moved eastward, and such scientists as Knaster, Orlicz, Saks
and Szpilrajn (Marczewski) arrived to Lwów. The Sovjets took eastern part of Poland with
Lwów and Wilno. They handed Wilno to Lithuania, which was for a while an independent
country, but soon joined the Sovjet Union. Thanks to this independence Antoni Zygmund
could emigrate to the United States, together with his family, and developped there later
a great school of analysis. Lwów was incorporated to Ukraine – a part of Sovjet Union -
but the University still worked with some Sovjet sciencists added. Banach even became
a dean and all Polish mathematicians got there positions, including those who escaped
from the German part of Poland. Also a volume of Studia Mathematica (almost ready
for publication before the war) was published in 1940; only the Ukrainian abstracts had
to be added. The meetings in Scottish Café continued (with a small break caused by
the war) and new problems appeared (some of them were given by P.S. Alexandrov and
S.L. Sobolev). In June 1941 Germany attacked Sovjet Union and took both Wilno and
Lwów. All openly done scientific and educational work stopped in all of Poland. However
some clandestine education was organized, mostly in Warsaw and in Kraków. It was very
dangerous both for pupils and teachers, but even in such circumstances some Ph.D. degrees
were granted).
Poland lost during the second world war about 50% of its mathematicians by death
or emigration. Still before the war some mathematicians, as Kac, Ulam, Tarski and an
excellent statistician Jerzy Splawa-Neyman emigrated. The main reason was a shortage
of positions. Before the World War II Poland had only 23 chairs of mathematics and 27
auxiliary positions, so that even such mathematicians as Nikodym or Orlicz had to teach in
secondary schools. The reconstructed in 1945 year volume of Fundamenta Mathematicae
(ready for print at the beginning of the war) contains a list of 22 names of mathematicians
who perished during the war (mostly in the concentration camps). This list contains
such names as Auerbach, Kaczmarz, Lindenbaum, Lomnicki, Marcinkiewicz, Rajchman,
Saks, Schauder and Schreier. Also Banach and Mazurkiewicz passed away just after the
end of war. In this situation the task of restoration of mathematics in Poland was very
hard. We lost not only many great mathematicians, but also one generation of youg
people, who either perished during the war, or were not sufficiently educated. Consequetly
there was a serious generation gap after the war. There were older mathematicians like
Sierpiński, Kuratowski, Borsuk and Mazur and younger collegues about thirty years old,
while the intermediate generation consisted only with a few mathematicians like Mostowski,
Marczewski (during the war he had to change his name from Szpilrajn) Mikusiński or
Sikorski. Also a five years break of research was very harmful. Here the situation was
quite different from that in Western Europe (I was very surprised when learning that, for
instance, in France or Netherlands the mathematicians worked more or less normally and
published their papers). One should add also the losses in literature (for instance, the
mathematical library of the Warsaw University was burned off in 1939, not mentioning the
lack of information on recent results). Also the change of borders, with Lwów University

7
moved to Wroclaw (former German Breslau), and Wilno University moved to Toruń with
all documentation and libraries left over, added to our loses. Nevertheless the job of
reconstruction and augmentation of Polish mathematics was done, but it should be the
subject of a different story.

References

[1] S. Banach, Sur les opérations dans les ensembles abstraits et leur application aux
équations integrales, Fund. Math. 3 (1922), 133-181.
[2] — , Sur le problème de la mesure, Fund. Math. 4 (1923), 7-33.
[3] — , Theory of Operations. vol.I. Linear Operations (in Polish), Warszawa 1931.
[4] — , Théorie des Opérations Linéaires, M.M. vol. 1, Warszawa 1932.
[5] S. Banach and A Tarski, Sur la décomposition des ensembles de points en parties
respectivement congruentes, Fund. Math. 6 (1924), 244-277.
[6] R. Kaluża, The Life of Stefan Banach, Birkhaüser, Boston 1996.
[7] K. Kuratowski, Topologie I, M.M. vol. 3, Warszawa 1933.
[8] — , A Half Century of Polish Mathematics, International series in pure and applied
mathematics vol. 108, Pergamon Press and PWN, 1980.
[9] R. D. Mauldin (ed.), Mathematics from the Scottish Café, Birkhaüser 1981.
[10] S. Saks, Theory of the Integral, M.M. vol. 7, Warszawa 1937.
[11] W. Sierpiński, Hypothèse du Continu, M.M. vol. 4, Warszawa 1934.
[12] H. Steinhaus, Stefan Banach , Studia Math. Ser. Spec. Z.1. (1963), 7-15.
[13] S. Ulam, Adventures of a Mathematician, University of California Press, Berkeley
1976.
[14] A. Zygmund, Trigonometrical Series, M.M. vol. 5, Warszawa 1935.

Mathematical Institute, Polish Academy of Sciences