Iterative square root? sqrt(2+sqrt(2+sqrt(

  • Thread starter Damidami
  • Start date
The other day I was playing with my calculator and noticed that

[tex]\sqrt{2+\sqrt{2+\sqrt{2+\sqrt{2+...}}}} \approx 2[/tex]

But, what is that kind of expression called? How does one justify that limit?
And, to what number exactly does converge, for example:

[tex]\sqrt{1+\sqrt{1+\sqrt{1+\sqrt{1+...}}}} \approx 1.6161[/tex]

[tex]\sqrt{3+\sqrt{3+\sqrt{3+\sqrt{3+...}}}} \approx 2.3027[/tex]

Any references where I could read about these subjects?

Another question. Considering real [tex]x>1[/tex], we have:
[tex]\Gamma(x) - x^1 = 0[/tex] then [tex] x \approx 2[/tex]

But how does one justify that? And what are the exact values of these functions:

[tex]\Gamma(x) - x^2 = 0[/tex] then [tex] x \approx 3.562382285390898[/tex]
[tex]\Gamma(x) - x^3 = 0[/tex] then [tex] x \approx 5.036722570588711[/tex]
[tex]\Gamma(x) - x^4 = 0[/tex] then [tex] x \approx 6.464468490129385[/tex]

They are called Nested Radicals. There are references in the link. I am unfortunately not familiar with their theory.
They are called Nested Radicals. There are references in the link. I am unfortunately not familiar with their theory.
Thank you slider142!
That answered my first question.
Does anyone know about my second question? Or any further references?
[tex]\sqrt{2+\sqrt{2+\sqrt{2+\sqrt{2+...}}}} = x[/tex]

For something like this, you can rewrite the equation as

[tex]\sqrt{2+x} = x[/tex]

And then the infinite equation is captured in a finite form. From there, you simply square both sides.

[tex]2+x = x^2[/tex]

And solve for x.

But I don't know much more than that! Don't forget that square-roots are non-negative.
For nested radicals of the form

[tex]\sqrt{a + \sqrt{a + \sqrt{a + \dots}}}[/tex]

using the trick

[tex]\sqrt{a + x} = x[/tex]

works very well. Two roots will emerge, but only one is positive (the other is extraneous).

In the case when a = 2, x = 2.

When a = 1, then x = [itex]\phi = \frac{1 + \sqrt{5}}{2} \approx 1.618034[/itex] a.k.a. the golden ratio.

As to the second question regarding the Gamma function, I'm not sure much theory is available.



Staff Emeritus
Science Advisor
Gold Member
For a dose of rigor -- we have to be sure the limit really exists before we can compute it with such tricks!

In this case, it's easy: the value is the limit of an increasing sequence, and limits of increasing, extended real number-valued sequences always exist.

It's important to notice that extended real numbers come into play here! The equation
2 + x = x²​
has three relevant solutions: -1, 2, and [itex]+\infty[/itex]. We know the limit exists, so it has to have one of those three values. It's easy to rule out -1, but more work is needed to decide between 2 and [itex]+\infty[/itex].
If we examine the sequence [itex]\{a_n\}_{n=0}^{\infty}[/itex] when [itex]a_0 = \sqrt{2}[/itex] and


Then it is possible to show by induction that [itex]a_n \leq 2 [/itex] for all n so the [itex]+\infty[/itex] case is impossible.

But you are correct, this possibility does need to be ruled out, Hurkyl.

Last edited:

Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving

Hot Threads