Disclaimer: I'm assuming your students are familiar with (partial) derivatives...
Let the ellipse be in standard position with foci at $F=(a, 0)$ and $F'=(-a, 0)$, and let $T=(x,y)$ be a general point on the ellipse. Then we have the invariance
$\|FT\| + \|F'T\| = \text{constant}$, that is
$$
f(x,y) = \text{constant}.
$$
where $f(x,y) := \|FT\| + \|F'T\| = \sqrt{(x-a)^2 + y^2} + \sqrt{(x+a)^2+y^2}$.
Let $(x,y) = (x(t), y(t))$ by any smooth paremetrization of the ellipse. For a tangent to the ellipse, we must have $\frac{d f(x,y)}{d t}=0$. For convenience, we will be using the following notation
- $\mathbf{\widehat{u}}:=\frac{x-a}{\|FT\|}\mathbf{\widehat{i}} + \frac{y}{\|FT\|}\mathbf{\widehat{j}}$ is the unit vector along $\overset{\longrightarrow}{FT}$
- $\mathbf{\widehat{v}}:=\frac{x+a}{\|FT\|}\mathbf{\widehat{i}} + \frac{y}{\|FT\|}\mathbf{\widehat{j}}$ is the unit vector along $\overset{\longrightarrow}{F'T}$
- $\mathbf{T}=\frac{\partial x(t)}{\partial t}\mathbf{\hat{i}} + \frac{\partial y(t)}{\partial t}\mathbf{\hat{j}}$ is the tangent at $T$.
Note that the perpendicular bisector of the angle $\measuredangle FTF'$ is the vector $\mathbf{\widehat{u}} + \mathbf{\widehat{v}}$.
Now, one computes $\frac{\partial f(x,y)}{\partial x} = \frac{x-a}{\sqrt{(x-a)^2 + y^2}}+\frac{x+a}{\sqrt{(x+a)^2 + y^2}}=\frac{x-a}{\|FM\|}+\frac{x+a}{\|F'M\|}$ and
$\frac{\partial f(x,y)}{\partial xy} = \frac{y}{\sqrt{(x-a)^2 + y^2}}+\frac{y}{\sqrt{(x+a)^2 + y^2}}=\frac{y}{\|FM\|}+\frac{y}{\|F'M\|}$. Thus
$$
\begin{split}
0 = \frac{d f(x,y)}{d t} &= \left(\frac{x-a}{\|FM\|} + \frac{x+a}{\|F'M\|}\right)\frac{\partial x}{\partial t} + \left(\frac{y}{\|FM\|} + \frac{y}{\|F'M\|}\right)\frac{\partial y}{\partial t}\\
&= (\mathbf{\widehat{u}} + \mathbf{\widehat{v}}).\textbf{T}
\end{split}
\tag{1}
$$
(1) says that the perpendicular bisector $\mathbf{\widehat{u}} + \mathbf{\widehat{v}}$ of the angle $\measuredangle FTF'$ is perpendicular to the tangent at $T$, and we are done. $\quad\quad\Box$
proof reflection ellipse. For instance, see "How to geometrically prove the focal property of ellipse?". – Blue Mar 28 '20 at 16:51