Let $g: \mathbb{R^2} \to \mathbb{R}$.
How can I prove that $g$ is continuous in its origin, but not totally differentiable?
If I take
$$g(\frac{1}{n},\frac{1}{n}) = \frac{1}{n\sqrt{2}} \to 0 \text{ for } n \to \infty $$
Or rather:
$$|x,y| \leq \frac{1}{2} (x^2 + y^2) $$
from which we can follow
$$|g(x,y)| \leq \frac{1}{2} \sqrt{x^2 + y^2}$$
which proves continuity of $g$ in the origin $(0,0)$.
But how can I show that this function is not total differentiable?
Can I do the following estimation?
$$|g(x,y) - 0| = |y| \cdot \frac{x \cdot y}{\sqrt{x^2 + y^2}} \leq |y| \cdot 1 = |y| $$
From which it follows, that
$$|y| \leq |\sqrt{x^2 + y^2}| = ||(x,y)|| $$

By polar coordinates as $r \to 0$
$$\frac{xy}{\sqrt{x^2 + y^2}}=r \cos \theta \sin \theta$$
we see that $f(x,y)$ is continuous but since $f_x(0,0)=f_y(0,0)=0$ by definition of differential we have
$$\lim_{(h,k)\to (0,0)} \frac{\frac{hk}{\sqrt{h^2 + k^2}}}{\sqrt{h^2 + k^2}}=\frac{hk}{h^2 + k^2}$$
which doesn't exist.
Refer also to the related