Statistical Inference — spotting the one fully-specified (simple) hypothesis

Let $X\sim N(\mu,\sigma^2)$. If both $\mu$ and $\sigma^2$ are unknown such that $-\infty<\mu<\infty,\sigma^2>0$, then which one of the following is not a composite hypothesis?