Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Square Root of Sum of Squares

  1. May 3, 2005 #1
    {{A_n}={\sqrt{\sum _{z=1}^{n}{z^2}}} } \\
    {{A_1}=1 } \\

    Is there a proof that only for n =1 or n=24 that An is an integer quantity?
  2. jcsd
  3. May 3, 2005 #2


    User Avatar
    Science Advisor
    Homework Helper

    [tex]A_n=\sqrt{\frac{n(n+1)(2n+1)}{6}}[/tex], and for n=24 you have 24/6 * 25 * 49.
  4. May 3, 2005 #3
    OK that is 4*25*49 all being squares proves [tex]A_{24}=2*5*7[/tex] and confirms what I have already said but that is no proof that another n could or does not exist such that [tex]A_n[/tex] is a integer quantity.
    Last edited: May 3, 2005
  5. May 3, 2005 #4
    I think there is a proof. I would check Dickson's History of the Theory of Numbers, if you are near a library that has it.
  6. May 3, 2005 #5
    I got the answer:

    The only numbers that are simultaneously square and pyramidal (the cannonball problem) are 1 and 4900, .........(Dickson 1952, p. 25; Ball and Coxeter 1987, p. 59; Ogilvy 1988), as conjectured by Lucas (1875, 1876) and proved by Watson (1918). The cannonball problem is equivalent to solving the Diophantine equation:

    [tex]y^2 = \frac{(x)(x+1)(2x+1)}{6}[/tex]

    Last edited: May 3, 2005
  7. May 4, 2005 #6
    Robert - thanks but is the fact that finding just one solution proof? it is just one answer. I have to find Watson's proof.

    Interesting to note is that

    [tex]\begin{array}{c}{{B_n}={\sqrt{\sum_{y=1}^{n}{\sum _{z=1}^{y}{z^2}}}}=\sqrt{\frac{n(n+1)^2(2n+1)}{6}} } \\{{B_1}=1 } \\{{B_{6}}=14\\{{B_{25}}=95}\\{etc}\\{B_{19785515999613069781581367686}=113006685964487096197915421007515390839216297500491081054\end{array}\[/tex]

    or solving the Diophantine equation

    [tex]y^2 = \frac{(x)(x+1)^2(2x+1)}{6}[/tex]

    has infinite number of integer solutions (the above shows the 1st, 2nd, 3rd and 50th solution) and are easy to generate numerically
    Last edited: May 4, 2005
  8. May 5, 2005 #7
    Sorry I made a mistake:
    [tex]{{B_n}={\sqrt{\sum_{y=1}^{n}{\sum _{z=1}^{y}{z^2}}}}=\sqrt{\frac{n(n+1)^2(n+2)}{12}} }[/tex]

    [tex]y^2 = \frac{(x)(x+1)^2(x+2)}{12}[/tex]

    The one factor is (n+2) and not (2n+1) and the divisor is 12 not 6

    Careless cut and paste on my part
    Last edited: May 5, 2005
  9. May 12, 2005 #8
    This has got me thinking. Is it possible to to put together squares with sides 1 to 24 to make a square of side 70. It looks unlikely as it is difficult to see how the small squares would fit in - for instance the 1 square has to be surrounded by larger squares. I wondered whether anyone has any more information on fitting squares together in this way.
  10. May 13, 2005 #9
    I doubt if you can fit 1 to 24 in a 70 square, very unlikely

    but 24 different integers squares can be packed into a square with 175 sides, solution in the attched file

    Attached Files:

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook