I'm reading a computation that shows that the mean value property for a function implies that function is harmonic. It includes the following computation:
\begin{align}\int_{B(x,r)}\Delta h(y) dy &= \int_{\partial B(x,r)} \nabla h(\vec{y})\cdot \vec{\nu}\text{ }dS(\vec{y}) \\ &= \int_{\partial B(x,r)} \nabla h(\vec{y})\cdot \frac{\vec{y}-\vec{x}}{r}\text{ }dS(\vec{y}) \\ &= \int_{\partial B(0,1)} \nabla h(\vec{x}+r\vec{z})\cdot \vec{z} \underbrace{\quad \quad}_{\text{why no factor of r?: } d\vec{y} = rd\vec{z}} \text{ }dS(\vec{z}) \\ &= \underbrace{\frac{d}{dr}}_{\text{why can this be puled through?}} \int_{\partial B(0,1)} h(\vec{x}+r\vec{z}) \text{ }dS(\vec{z}) \end{align}
Can anyone answer the questions posed in lines 3 and 4?