For the mean vector of a p-variate normal distribution ($p{geq}3$), the optimal estimation within the class of James-Stein type decision rules under the quadratic loss are given when the underlying distribution is that of a variance mixture of normals and when the norm ${parallel}underline{{ heta}}{parallel}$ in known. It also demonstrated that the optimal estimation within the class of Lindley type decision rules under the same loss when the underlying distribution is the previous type and the norm ${parallel}{ heta}-overline{ heta}underline{1}{parallel}$ with $overline{ heta}=frac{1}{p}sumlimits_{i=1}^{n}{ heta}_i$ and $underline{1}=(1,{cdots},1)^{prime}$ is known.