Contracting the boundary of a Riemannian 2-disc (Q897001)

In this article, the authors offer an answer to a question raised by M. Gromov and restated by \textit{S. Frankel} and \textit{M. Katz} [Ann. Inst. Fourier 43, No. 2, 503--507 (1993; Zbl 0780.53035)]. The question regards the possibility to give an upper estimate for the length of the family of curves used to homotopically contract the boundary of a Riemannian 2-disk into a point. This upper estimate is obtained using a certain constant, the length and the diameter of the disk, but also the area of the disk (this later ingredient was suggested by S. Frankel and M. Katz). The main tool is a method of dividing the disk into small portions (the Besicovitch lemma) which is applied inductively. The authors also study the length of closed curves sweeping-out a Riemannian 2-sphere, obtaining interesting results concerning the diastole of such closed manifolds.