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

In base 10, the sequence 49,4489,444889,... consists of all perfect squares.

Is this true for any other bases (greater than 10, of course)?

share|cite|improve this question
In base $b$, try looking at $\lfloor 2b^k/3 \rfloor^2$, where $\lfloor\cdot\rfloor$ is the floor function i.e. rounds numbers down to the nearest integer below. – anon Aug 3 '11 at 19:11
@anon: this series is $((2\cdot10^k+1)/3)^2$, so you should use the ceiling – Ross Millikan Aug 3 '11 at 19:32
Note, there is a post on the problem (in base $10$), here. – JavaMan Aug 3 '11 at 19:33
@Ross: You're correct, I don't know why I said floor. – anon Aug 3 '11 at 19:40
1331 is a cube in any base $b\geq 4$. – Christian Blatter Aug 3 '11 at 20:28
up vote 5 down vote accepted

No, it isn't.

The $n$'th term of your sequence in base $b$, if I understand correctly, is $1 + 8 \sum_{j=0}^{n-1} b^j + 4 \sum_{j=n}^{2n-1} b^j$. Consider the case $n=1$: $a_1 = 4 b + 9$. If that is a square, say $(2k+1)^2$ (since it is odd), we have $b = ((2k+1)^2 - 9)/4 = k^2 + k - 2$.
Now taking $n=3$: $a_3 = 4\,{k}^{10}+20\,{k}^{9}+4\,{k}^{8}-104\,{k}^{7}-64\,{k}^{6}+256\,{k}^{ 5}+120\,{k}^{4}-348\,{k}^{3}-24\,{k}^{2}+216\,k-71$. Consider $q = 2\,{k}^{5}+5\,{k}^{4}-{\frac {21}{4}}\,{k}^{3}-{\frac {103}{8}}\,{k}^{ 2}+{\frac {595}{64}}\,k+{\frac {891}{128}}$ (which comes from the Laurent series of $\sqrt{a_3/k^{10}}$ at $k=\infty$). It turns out that $a_3 - q^2 < 0$ for all real $k$, while $a_3 - (q - 1/128)^2 > 0$ for $k > 567.75$. Thus if $k$ is an integer $> 567$, $\sqrt{a_3}$ is between two numbers in of the form $\frac{\text{integer}}{128}$, and in particular is not an integer. Trying values of $k$ from 2 to 567, the only one that makes $a_3$ a square is $k=3$ (corresponding to $b=10$).

share|cite|improve this answer
Impressive! Thanks... – EagerLearner Aug 4 '11 at 18:39

More generally, if $b = 9 m + 1$ and $r = 4 m$, the corresponding sequence $a_n = 1 + 2 r \sum_{j=0}^{n-1} b^j + r \sum_{j=n}^{2n-1} b^j$ consists of squares, namely $a_n = \left( \frac{2(9m+1)^n+1}{3} \right)^2$.

share|cite|improve this answer

$((2*19^5+1)/3^2)=2724919437289,\ \ $ which converted to base $19$ is $88888GGGGH$. It doesn't work in base $13$ or $16$. In base $28$ it gives $CCCCCOOOOP$, where those are capital oh's (worth $24$).

This is because if we express $\frac{1}{9}$ in base $9a+1$, it is $0.aaaa\ldots$. So $\left (\frac{2(9a+1)^5+1}{3}\right)^2=\frac{4(9a+1)^10+4(9a+1)^5+1}{9}=$

$ (4a)(4a)(4a)(4a)(4a)(4a)(8a)(8a)(8a)(8a)(8a+1)_{9a+1}$

where the parentheses represent a single digit and changing the exponent from $5$ changes the length of the strings in the obvious way.

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.