In Griffiths E&M 4th edition, equation 10.28 reads $$\nabla \rho = \dot{\rho}\nabla t_r$$ This is in the context of computing the gradient of the retarded potential $$V(\vec{r'},tr)=\frac{1}{4\pi \epsilon_0}\int \frac{\rho(\vec{r'},t_r)}{\mathit{r}_1}d\tau'$$ where $$t_r \equiv t-\frac{r}{c}\\ \mathit{r}_1 \equiv |\vec{r}-\vec{r'}|.$$
The step is clear, using the product rule for the gradient $$\nabla V= \frac{1}{4\pi \epsilon_0}\int\left[\left(\nabla \rho \right )\frac{1}{\mathit{r}_1} + \rho \nabla \left(\frac{1}{\mathit{r}_1} \right ) \right ]d\tau'.$$
Then the problem I am having is understanding the next step. Without explanation or reference, states that $$\nabla \rho = \dot{\rho}\nabla t_r.$$
I can follow the rest of it, but I don't see how he goes from $\nabla \rho(r,t_r)$ to $\dot{\rho}\nabla t_r$.
It seems like a small matter in such a complex subject, but I could use some help explaining this. BTW I get that $\dot{\rho}$ is the time derivative of $\rho$.