Root functions, eigenvectors, associated vectors and the inverse of a holomorphic operator function (Q1076307)

For Banach spaces E and F, \(\phi(E,F)\) denotes the set of all Fredholm operators on E to F; for an open set U of \({\mathbb{C}}\), \(H(U,E)\) is the set of all holomorphic functions on U to E and \(H(U,E_ 0)=\{f\in H(U,E):f(U)\subset E_ 0\subset E\}\); for \(T\in H(G,\phi (E,F))\), \(Y\in H(G,E)\), where G is a region in \({\mathbb{C}}\), is called a root function of T, at \(\mu\in G\) if \(Y(\mu)\neq 0\) and \((TY)(\mu)=0\). Suppose T is a holomorphic Fredholm operator function on a domain G in \({\mathbb{C}}\) and that the inverse of T exists at least at one point. In this set up an earlier theorem of \textit{M. V. Keldysh} [Usp. Mat. Nauk 26, No.4(160), 15- 41 (1971; Zbl 0225.47008)] states that for T of the type \(T(\lambda)=I+K(\lambda)\), where \(K(\lambda)\) is compact for \(\lambda\), the inverse is a meromorphic function with poles exactly at the eigenvalues of T and the theorem also gives a representation of the operators in the singular part of \(T^{-1}\). In the present paper the authors present the theorem of Keldysh in the framework of root functions and exhibit a proof of it in this framework.