Return to my Mathematics pages
Go to my home page


Perfect Numbers

© Copyright 2001, Jim Loy

6 and 28 are called Perfect Numbers. The proper divisors (the divisors of a number, not including the number itself) of 6 are 1, 2, and 3, and 6=1+2+3. Similarly, the proper divisors of 28 are 1, 2, 4, 7, and 14 and 28=1+2+4+7+14. Are there any other perfect numbers, numbers equal to the sum of their proper divisors? Euclid (in Book IX, Proposition 36) actually showed that if p is prime, and if (2^p)-1 is also prime, then (2^(p-1))((2^p)-1) is perfect (please forgive all the parentheses, but my notation is limited by HTML. 2^p means 2 to the p power). Euler showed that all even perfect numbers are of this form.

Primes of the form (2^p)-1 are called Mersenne Primes. A number of the form (2^n)-1 cannot be a prime unless n is a prime. The nth Mersenne number is Mn=M(n)=(2^n)-1. Many people call M(n) a Mersenne number only if n is prime. The first few Mersenne primes are 3, 7, 31, 127, 8191, etc., which are primes. So the first few perfect numbers are 6, 28, 496, 8128, 33550336, etc. It was quickly shown that M(11), M(23), M(83), M(131) and others were not prime, even though n is prime in those cases. Mersenne apparently conjectured that M(n) was prime for n=2, 3, 5, 7, 13, 17, 19, 31, 67, 127, and 257, and that M(n) was not prime for any other n below 257. He was shown to be wrong for n=61, 67, 89, 107, and 257. It is relatively easy to test to see if a Mersenne number is prime, and so the largest known prime is often a Mersenne prime.

Here are the 37 known Mersenne primes and perfect numbers (from MathWorld):

p M(p) Perfect(p) year source
2 3 6    
3 7 28    
5 31 496    
7 127 8128    
13 8191 33550336 1461 later mentioned by Reguis and Cataldi
17 131071 8589869056 1588 Cataldi
19 524287 137438691328 1588 Cataldi
31 10 digits 2305843008139952128 1750 Euler
61 19 digits   1883 Pervouchine
89 27 digits   1911 Powers
107 33 digits   1913 Powers
127 39 digits   1876 Lucas
521 157 digits   1952 Lehmer
607 183 digits   1952 Lehmer
1279 386 digits   1952 Lehmer
2203 664 digits   1952 Lehmer
2281 687 digits   1952 Lehmer
3217 969 digits   1957 Riesel
4253 1281 digits   1961 Hurwitz
4423 1332 digits   1961 Hurwitz
9689 2917 digits   1963 Gillies
9941 2993 digits   1963 Gillies
11213 3376 digits   1963 Gillies
19937 6002 digits   1971 Tuckerman
21701 6533 digits   1978 Noll and Nickel
23209 6987 digits   1979 Noll
44497 13395 digits   1979 Nelson and Slowinski
86243 25962 digits   1982 Slowinski
110503 33265 digits   1988 Colquitt and Welsh
132049 39751 digits   1983 Slowinski
216091 65050 digits   1985 Slowinski
756839 227832 digits   1992 Gage and Slowinski
859433 258716 digits   1994 Gage and Slowinski
1257787 378632 digits   1996 Slowinski and Gage
1398269 420921 digits   1996 Armengaud, Woltman, et al.
2976221 895832 digits   1997 Spence
3021377 909526 digits   1998 Clarkson, Woltman, et al.

There may be unknown Mersenne primes less than the last two in the table, as that region has not been searched exhaustively.

It has been conjectured that there are infinitely many Mersenne primes, and therefore infinitely many perfect numbers. No odd perfect number has been found, up to 10^300. The educated guess is that there are no odd perfect numbers.


Addendum:

The book Mystifying Math Puzzles, by Steve Ryan, has this puzzle, labeled "The Perfect Pyramid":

?
28
496
8128
130,816
2,096,128

"What is the missing number that belongs on top of this pyramid, and begins this progressively growing logical number secquence?" Some of the numbers should look familiar. The answer in the back of the book says this:

The missing number is six. All of the numbers on this pyramid are "perfect." A perfect number is a number whose total divisions [proper divisors] equal itself. Ex: 1,2,3 are all divisors of 6 and total six. 1,2,4,7,14 are all divisors of 28 and total 28. There are only six perfect numbers between 1 and 3,000,000.

Of course two of the above numbers are not perfect numbers. And there are only four perfect numbers between 1 and 3,000,000. 130,816 is 2^8(2^9-1), and fails to be a perfect number on two counts: 9 is not a prime, and so 2^9-1 is also not a prime. 11 is a prime, but 2,096,128 (2^10(2^11-1)) fails to be a perfect number because 2^11-1 is not a prime.


Return to my Mathematics pages
Go to my home page