On the Milnor \(K\)-theory of the function fields of quasihomogeneous cones (Q1908236)

Let \(v\) be a discrete valuation on a field \(F\) with residue field \(F_v\). In ``Algebraic \(K\)-theory and quadratic forms'' [Invent. Math. 9, 318-344 (1970; Zbl 0199.55501)] \textit{J. Milnor} showed that -- in analogy with the theory of Witt rings -- there are two induced residue homomorphism on higher Milnor \(K\)-theory \[ \psi_v : K_n (F) \to K_n (F_v) \] and \[ \partial_v : K_n (F) \to K_{n - 1} (F_v). \] Assume now that \(V\) is a normal quasi-homogeneous variety, hence a cone over a normal projective variety \(C\), with function field \(F(V)\). The author describes the kernel and the cokernel of the direct sum of all the residue homomorphisms on \(K_* (F(V))\) in terms of the \(K\)-theory of the function field of \(C\) and the \(K\)-groups associated with subvarieties of \(V\) of codimension 2. For example, if \(V = \text{Spec} F[x,y]\), the author shows that there is a homomorphism \(\delta\) and an exact sequence \[ 0 \to K_n (F) @>i^*>> K_n \bigl( F(x,y) \bigr) @>\partial>> \bigoplus_{\text{ht} p = 1} K_{n - 1} (F_p) @>\delta>> \bigoplus_{\text{ht} p = 2} K_{n - 2} (F_p) \to 0, \] where \(i^*\) is induced from the canonical map \(F \to F (x,y)\) and \(\partial\) is a sum of second residue homomorphisms.