# Borsuk problem

One of the fundamental problems in combinatorial geometry: Is it possible, for any bounded set of diameter $d>0$ in an $n$-dimensional Euclidean space, to make a decomposition into not more than $n+1$ subsets with diameters smaller than $d$? The problem was formulated by K. Borsuk  who noted that it was not possible to subdivide an $n$-dimensional simplex and an $n$-dimensional sphere in $\mathbf R^n$ into $n$ parts of a smaller diameter. The problem has a positive solution for $n=2,3$, but only partial results have been obtained for $n>3$. Thus, for instance, the problem has been positively solved for any bounded smooth convex set in $\mathbf R^n$ . It has been proved that the solution of Borsuk's problem can be reduced to the case of sets of constant width. If $a(F)$ is the smallest number of parts of a diameter smaller than $d$ into which a set $F\subset\mathbf R^n$ can be subdivided, then the equality $a(F)=3$ is valid for a figure $F\subset\mathbf R^2$ of diameter $d$ if and only if $\mathbf R^2$ contains a unique figure of constant width $d$ containing $F$ . This fact cannot be directly generalized to the case $n>2$. The Borsuk problem is closely related to the illumination problem and to the Hadwiger hypothesis, which is a generalization of the Borsuk problem in which $\mathbf R^n$ is replaced by a finite-dimensional normed space.