Integers whose Squares end in 444
Theorem
The sequence of positive integers whose square ends in $444$ begins:
- $38, 462, 538, 962, 1038, 1462, 1538, 1962, 2038, 2462, 2538, 2962, 3038, 3462, \ldots$
This sequence is A039685 in the On-Line Encyclopedia of Integer Sequences (N. J. A. Sloane (Ed.), 2008).
Proof
\(\ds 38^2\) | \(=\) | \(\ds 1444\) | ||||||||||||
\(\ds 462^2\) | \(=\) | \(\ds 213 \, 444\) | ||||||||||||
\(\ds 538^2\) | \(=\) | \(\ds 289 \, 444\) | ||||||||||||
\(\ds 962^2\) | \(=\) | \(\ds 925 \, 444\) | ||||||||||||
\(\ds 1038^2\) | \(=\) | \(\ds 1 \, 077 \, 444\) | ||||||||||||
\(\ds 1462^2\) | \(=\) | \(\ds 2 \, 137 \, 444\) | ||||||||||||
\(\ds 1538^2\) | \(=\) | \(\ds 2 \, 365 \, 444\) | ||||||||||||
\(\ds 1962^2\) | \(=\) | \(\ds 3 \, 849 \, 444\) | ||||||||||||
\(\ds 2038^2\) | \(=\) | \(\ds 4 \, 153 \, 444\) | ||||||||||||
\(\ds 2462^2\) | \(=\) | \(\ds 6 \, 061 \, 444\) | ||||||||||||
\(\ds 2538^2\) | \(=\) | \(\ds 6 \, 441 \, 444\) | ||||||||||||
\(\ds 2962^2\) | \(=\) | \(\ds 8 \, 773 \, 444\) | ||||||||||||
\(\ds 3038^2\) | \(=\) | \(\ds 9 \, 229 \, 444\) | ||||||||||||
\(\ds 3462^2\) | \(=\) | \(\ds 11 \, 985 \, 444\) |
All such $n$ are of the form $500 m + 38$ or $500 m - 38$:
\(\ds \paren {500 m + 38}^2\) | \(=\) | \(\ds 250 \, 000 m^2 + 38 \, 000 m + 1444\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \paren {250 m^2 + 38 m + 1} + 444\) |
\(\ds \paren {500 m - 38}^2\) | \(=\) | \(\ds 250 \, 000 m^2 - 38 \, 000 m + 1444\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \paren {250 m^2 - 38 m + 1} + 444\) |
and it is seen that all such numbers end in $444$.
Now we show that all such numbers are so expressed.
In Squares Ending in Repeated Digits, we have shown the only numbers with squares ending in $44$ ends in:
- $12, 38, 62, 88$
hence any number with square ending in $444$ must also end in those numbers.
Suppose $\sqbrk {axy}^2 \equiv 444 \pmod {1000}$, where $a < 10$ and $\sqbrk {xy}$ is in the above list.
For $\sqbrk {xy} = 12$:
\(\ds 444\) | \(=\) | \(\ds \paren {100 a + 12}^2\) | \(\ds \pmod {200}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 10000 a^2 + 2400 a + 144\) | \(\ds \pmod {200}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 144\) | \(\ds \pmod {200}\) |
This is a contradiction.
Similarly for $\sqbrk {xy} = 88$:
\(\ds 444\) | \(=\) | \(\ds \paren {100 a + 88}^2\) | \(\ds \pmod {200}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 10000 a^2 + 17600 a + 7744\) | \(\ds \pmod {200}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 144\) | \(\ds \pmod {200}\) |
Again, a contradiction.
For $\sqbrk {xy} = 38$:
\(\ds 444\) | \(=\) | \(\ds \paren {100 a + 38}^2\) | \(\ds \pmod {1000}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 10000 a^2 + 7600 a + 1444\) | \(\ds \pmod {1000}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 600 a + 444\) | \(\ds \pmod {1000}\) |
The solutions to $600 a \equiv 0 \pmod {1000}$ are $a = 0$ or $5$.
Hence:
- $\paren {500 n + 38}^2 \equiv 444 \pmod {1000}$
Similarly, for $\sqbrk {xy} = 62$:
\(\ds 444\) | \(=\) | \(\ds \paren {100 a + 62}^2\) | \(\ds \pmod {1000}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 10000 a^2 + 12400 a + 3844\) | \(\ds \pmod {1000}\) | |||||||||||
\(\ds \) | \(=\) | \(\ds 400 a + 844\) | \(\ds \pmod {1000}\) |
The solutions to $400 a + 400 \equiv 0 \pmod {1000}$ are $a = 4$ or $9$.
Hence:
- $\paren {500 n + 462}^2 \equiv \paren {500 n - 38}^2 \equiv 444 \pmod {1000}$
$\blacksquare$
Sources
- 1964: Albert H. Beiler: Recreations in the Theory of Numbers
- 1986: David Wells: Curious and Interesting Numbers ... (previous) ... (next): $462$
- 1986: David Wells: Curious and Interesting Numbers ... (previous) ... (next): $1444$
- 1997: David Wells: Curious and Interesting Numbers (2nd ed.) ... (previous) ... (next): $462$
- 1997: David Wells: Curious and Interesting Numbers (2nd ed.) ... (previous) ... (next): $1444$