Chernyi's series method is not proper for the case that(γ-l)/(γ+l)<<2/(γ+1)(M2sin2β)(γ=cp/cv-adiabatic index number, M-Much number, shock incidence). In this paper, we only suppose that in the neighbour of the shock, there exists a shock layer in which the density of the gas is very big, but we do not remove the case that (γ-1)/(γ+1)<<2/(γ+1)(M2sin2β).