Why is it wrong to think that $$\lim_{x \to 0} (1+\tan x)^\frac{1}{x} = (1+0)^\frac{1}{x}=1^\infty=1$$ because $\lim_{x \to 0} \tan x$ and $\lim_{x \to 0} \frac{1}{x}=\infty$? I already know the correct answer but could not figure out why this way is not correct.
-
2Related: "Why is $1^\infty$ considered to be an indeterminate form?" – Blue Aug 05 '16 at 09:09
-
it's the 1/x that is the indeterminate form, what you have is a limit where the exponent is tending towards infinity, but the case where (1 + 0) is multiplied by itself infinite times never actually happens, because x = 0 can never happen in that equation - if you consider (1 + ax) ^(1/x) your method of 1^infinity goes wrong - try it! – Cato Aug 05 '16 at 09:53
2 Answers
$1^\infty$ is indeterminate form, you can't say it is $1$.

- 6,475
-
I'd say it is the 1/x that is the indeterminate form, you can never plug zero into it. Because you can't ever put zero into your 1/x, you never get (1 + 0) to raise to a power - you always have (1 + something)^(something massive). But as far as I can see 1^x = 1 as x tends to infinity. – Cato Aug 05 '16 at 09:57
What you are trying to do is to "split up" the limit taking procedure. In your case you to take the limit of the base which is $1$ and you then want to take the limit of the exponent, which you say is $\infty$.
You are probably familiar with this precedure from expressions like these:
\begin{equation} \lim_{x \to 0} (f(x) \cdot g(x)) = (\lim_{x \to 0} f(x)) \cdot (\lim_{x \to 0} g(x)) \end{equation}
Recall however that this only has to work when both limits of $f$ and $g$ exist as real numbers.
The very same splitting will usually work for most "nice" operations, like in your case exponentiation. (If you hit the limit $0^0$ however, it will not!)
But since $\infty$ is not a real number, it is (here) just a symbol denoting unbounded sequences, there is no reason the splitting should work.
$1^{\mathrm{bla}}$ when bla is no real number does not really make any sense at all.

- 1,167