The definition of $\lim_{x\to 0} f(x) = \infty$ is:
$\forall N\in\mathbb{R}\;\exists \delta\gt 0\; (0<\lvert x\rvert<\delta\implies f(x)>N)$
The definition of "$\lim_{x\to 0} f(x)$ does not exist" is:
$\forall L\in\mathbb{R}\; \exists \epsilon\gt 0\; \forall \delta\gt 0 \; \exists x\in \mathbb{R}\; (0<\lvert x\rvert<\delta\; and\; \lvert f(x)-L \rvert\gt\epsilon)$
The question makes sense intuitively and geometrically, but how do I prove directly from the definitions that the first definition implies the second?
Thank you!
Suppose you had $\lim_{x\to 0} f(x) = L < \infty$ and $\lim_{x\to 0} f(x) = \infty$. Then from the first, applying it with $\epsilon := 1$, you can find $\delta > 0$ such that $|f(x) - L| < 1$ whenever $0 < |x| < \delta$. Similarly, from the second, you can find $\delta' > 0$ such that $f(x) > L + 1$ whenever $0 < |x| < \delta'$.
Now, can you see from here how to reach a contradiction?