No label defined (Q4359655)

The authors offer a variant of \textit{L. Vietoris}' argument [J. Reine Angew. Math. 186, 1-15 (1944; Zbl 0060.19108)] deriving \[ f(x+y)=f(x)f(y)-g(x)g(y), \] for real valued functions \(f,g\) of a real variable from \[ f(x-y)=f(x)f(y)+g(x)g(y) \] and then claims to show that from these two functional equations the following two follow: \[ g(x+y)=g(x)f(y)+f(x)g(y),\;g(x-y)=g(x)f(y)-f(x)g(y). \] The latter argument seems to be unfinished and has, in the reviewer's opinion, the flaw that it assumes that \(\cos\theta(x+y)=\cos(\theta(x)+\theta(y))\) implies either \(\theta(x+y)=\theta(x)+\theta(y)\pmod{2\pi}\) everywhere or \(\theta(x+y)=-\theta(x)-\theta(y)\pmod{2\pi}\) everywhere, while for noncontinuous \(\theta\) it could imply the first congruence for some \((x,y)\), the second for others. About the above four functional equations and connections among them much has been written both before that paper of Vietoris and after, e.g. by \textit{W. H. Wilson} [Bull. Am. Math. Soc. 26, 300-312 (1919-20; JFM 47.0320.01)], \textit{E. Vincze} [Publ. Math. 9, 149-163, 314-323 (1962; Zbl 0125.35803)], \textit{J. K Chung, Pl. Kannappan} and \textit{C. T. Ng} [Linear Algebra Appl. 66, 259-277 (1985; Zbl 0564.39002)]. For a survey see \textit{J. Aczél} and \textit{J. Dhombres} [Functional equations in several variables (1989; Zbl 0685.39006), pp. 219-227].