Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

My calculus book says that with: $$a=x^{x^{x^{.^{.^{.}}}}}$$ (exponent tower goes on forever), then: $$x=a^\frac{1}{a}$$

I tried it out with $a=3$ so $x=3^\frac{1}{3}$ and then ran a python program to test it. I did: $$(3^\frac{1}{3})^{(3^\frac{1}{3})^{(3^\frac{1}{3})^{.^{.^{.}}}}}$$

Between 500 runs and 10.000 runs the answer stayed: 2.4780526802882967

Maybe I don't appreciate enough what infinity means, but why isn't it closer to $3$ after 10.000 runs? Is the answer $x=a^\frac{1}{a}$ really correct? Will this really finally be 3 if it would run infinitely?

For reference, this was the little python program:

for i in range(0,100000):
share|cite|improve this question
What exactly does your calculus book say? I somehow doubt that the infinite exponent tower is defined for $\lvert x\rvert > 1$. – Clement C. Jan 31 at 18:10
@ClementC.: You would think so, wouldn't you? But your (and my) intuition is faulty $-$ the sequence of finite exponent towers is bounded above by $a$. – TonyK Jan 31 at 18:38
@ClementC. Such power tower only converges for $e^{-e} \leq x \leq e^{\frac1e}$. If it converges, then it must statisfy $x^a=a$ and hence $x=a^{\frac1a}$. – wythagoras Jan 31 at 18:38
My bad. But then, if I understood correctly this and the answer below (I feel I am missing something)-- the issue is that the sequence defined by $x_1 = x = 3^{1/3}$ and $x_{n+1} = x_n^x$ does converge, but not to the tower in question? @wythagoras (i.e., it converges to one of the two values $\ell$ such that $\frac{\ln \ell}{\ell} = \frac{\ln 3}{3}$, but "sadly" not the right one.) – Clement C. Jan 31 at 18:55
@wythagoras So basically the issue is that the function doesn't converge for $x=3^\frac{1}{3}$? – Imre Végh Jan 31 at 20:26
up vote 12 down vote accepted

You are confusing the direct and converse implication.

The book doesn't say that if we set $x=a^{\frac{1}{a}}$ the sequence converges.

It says that if the series converges to $b$, then you have $x=b^{\frac{1}{b}}$, where $b$ is the limit.

If you apply the theorem in this case, it tells you that your $x$ is equal to the limit to the limit, not that

If everything worked well in your programming, this means that you must have $$3^{\frac{1}{3}}= 2.4780526802882967^{\frac{1}{ 2.4780526802882967}} $$

Note that in general, a number $x$ could be written in more than one way as $a^\frac{1}{a}$, but the tower can only converge to one $a$. This means that if you start with another one, the limit will be different.

share|cite|improve this answer

$$y_{n+1}=x^{y_n}$$ $$\ln y_{n+1}=y_n \ln x=y_n (\frac{1}{3}\ln 3)$$ What you are getting is $y_{n+1}=y_n=2.4780526802882967$,

$$\frac{\ln y_{n}}{y_n}= \frac{1}{3}\ln 3$$

Which should have been true only for $y_n =3$ if $\frac{\ln x}{x}$ was a one to one function except that it's not.Plot

The above plot shows the function $\frac{\ln x}{x}$ .Observe that the line $y=\frac{\ln 3}{3}$ intercepts the graph at 2 points- $x=3$ and $x=2.4780526802882967$.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.