How can I use $T=\left(\frac{\partial E}{\partial S}\right)_V$ and $P=-\left(\frac{\partial E}{\partial V}\right)_S$ to derive
$$\left(\frac{\partial T}{\partial V}\right)_S = -\left(\frac{\partial P}{\partial S}\right)_V$$
The book (Statistical Physics by F. Mandl) suggests that $\frac{\partial^2 E}{\partial V \partial S} = \frac{\partial^2 E}{\partial S \partial V}$, but I can't make the connection as to how this helps. He calls this "cross-differentiation", i.e. to form the two derivatives given.
However, how can his suggestion be true? If we take the 2nd derivative of $E$ using the equation for $T$ and $P$ given above, then
$$\partial T=\left(\frac{\partial^2 E}{\partial S^2}\right)$$
$$\partial P=-\left(\frac{\partial^2 E}{\partial V^2}\right)$$