Sums of Consecutive Integers
   

   

 Math Help -> Puzzles -> Sums of consecutive integers -> Answer 

Sums of Consecutive Integers

In what ways can 1,000,000 be expressed as the sum of consecutive positive whole numbers?

Source: Durango Herald Puzzle Corner, by Dick Gibbs

Solution

The answer is 7 ways:

The 640 numbers 1243+1244+...+1882=1,000,000,
the 625 numbers 1288+1289+...+1912=1,000,000
the 128 numbers 7749+7750+...+7876=1,000,000
the 125 numbers 7938+7939+...+8062=1,000,000
the 25 numbers 39,988+39,989+...+40,012=1,000,000
the 5 numbers 199,998+199,999+...+200,002=1,000,000
and the 1 number 1,000,000.

Some people have raised a concern that there is no such thing as one consecutive number.  That is, the single number 1,000,000 is not itself an expression of the sum of consecutive numbers.  They point to the use of plural in the question to say that at the very least, the question is deceptive in its ambiguity.  This is a fair concern, and perhaps the writers of the puzzle intended this ambiguity as an extra degree of difficulty, frustrating as that may be for some people!

An informal consensus of mathematicians is that a property, such as "consecutive", is satisfied as long as no element of the sequence violates that property.  Thus, a sequence of one term has the property of its terms being "consecutive" because there is no non-consecutive term.  For that matter, the sequence of zero terms can be said to be consecutive for the same reason.

Ruth, a participant in the "Ask NRICH" discussion group, pointed out that the question might also have been posed,

In what ways can 1,000,000 be expressed as the difference of two triangular numbers?

A more general problem

What math enthusiast can pass up the opportunity to pluck a general rule from a specific instance?  Here's a way to express a more general version of the problem, the solution to which sheds light on this problem...

In what ways can N be expressed as the sum of one or more consecutive positive integers?

If mn=N, with n odd, then n consecutive positive numbers have a "middle number", which is the average of the n numbers.  If the sum of the n numbers is N, then this average is m.

Example: 3*7=21, so let m be 3 and n be the odd number of terms, 7.  Then we can find 7 consecutive integers whose average is 3, and therefore the sum of the 7 numbers is 21.  These numbers are 0, 1, 2, 3, 4, 5, and 6.  Oops!  These aren't all positive integers, so this isn't a solution to the question.

Now let 7*3=21, with m=7 the average of the terms, and n=3 is the number of terms.  Then we have 6+7+8=21, which is a solution to the question.  A little thought will convince you that an odd factor of N can be used to make an odd number, n, of terms whose average is m and whose sum is N only if 2m > n.  This is true because the smallest term is m-n/2+1/2, which must be greater than 1/2.

Let me use this thinking to refine my statement about sums of an odd number of consecutive terms:

If mn=N, with n odd, and 2m > n, then the n consecutive positive numbers ranging from m-n/2+1/2 to m+n/2-1/2, has an average of m, and a sum of N.

What about an even number of consecutive positive numbers?  Can they add up to N?  Let's see.  An even number of consecutive numbers has no "middle" number.  The middle, if you can call it that, of the numbers is the average of the middle two.  It's also the average of the first and last.  And it's the average of the second and second-to-last.  I'm sure you get the idea: the average of an even number of consecutive integers is a number midway between two integers, which we can call n/2, where n is an odd number.

Let 2m be this even number of consecutive integers whose average is n/2.  Then, N, the sum of the integers is given by N=(2m)(n/2)=mn.

Example: Just for fun, let's use the same example as before: 3*7=21, letting m be 3 and n be 7.  So we can have 2m (six) terms whose average is n/2 (3 1/2) that add up to 21.  They are 1+2+3+4+5+6=21.

Now, let 7*3=21, with m=7 (so there will be 2m=14 terms) and n=3 (so the average of the terms is n/2=3/2).  These terms are -5+-4+-3+-2+-1+0+1+2+3+4+5+6+7+8=21.  Oops!  Once again, we have a sequence that doesn't satisfy the problem because some of the terms are not positive integers.  In order for all the terms to be positive, the smallest term, which is n/2-m+1/2, must be greater than 1/2, so n > 2m.

As before, I will refine this second statement about the sums of an even number of consecutive terms:

If mn=N, with n odd, and n > 2m, then the 2m consecutive positive numbers ranging from n/2-m+1/2 to n/2+m-1/2, has an average of n/2, and a sum of N.

Together, the two statements in the boxes, above, give a solution for each odd divisor, n, of N.  Each solution is a unique set of consecutive positive integers whose sum is N.

Are there any other sequences of consecutive positive integers that sum to N?  No, because every sequence either has an even number of terms, in which case the number of terms is 2m for some number, m, and their average is n/2 for some odd number, n, so their sum is mn; --OR-- it has an odd number of terms, in which case the number of terms is some odd number, n, and their average is an integer, m, so their sum is mn.  In either of these cases, every sequence summing to the product mn is identified by one of the boxed statements, above.

If N=2kM, where M is odd, then the odd divisors of n is the same as divisors of M.  The number of such divisors can be obtained from the prime factorization of M, by taking the product of the numbers one greater than each exponent.

Example: Let's use the boxed statements to find the 7 sequences whose sum is 1,000,000.  The odd factors of 1,000,000 are 1, 5, 25, 125, 625, 3125, and 15625.

When n=15625, m=64, n > 2m, so 2*64 numbers averaging 15625/2 sum to: 7749+7750+...+7876=1,000,000
When n=3125, m=320, n > 2m, so 2*320 numbers averaging 3125/2 sum to: 1243+1244+...+1882=1,000,000
When n=625, m=1600, 2m > n, so 625 numbers averaging 1600 sum to: 1288+1289+...+1912=1,000,000
When n=125, m=8000, 2m > n, so 125 numbers averaging 8000 sum to: 7938+7939+...+8062=1,000,000
When n=25, m=40,000, 2m > n, so 25 numbers averaging 40,000 sum to: 39,988+39,989+...+40,012=1,000,000
When n=5, m=200,000, 2m > n, so 5 numbers averaging 200,000 sum to: 199,998+199,999+...+200,002=1,000,000
When n=1, m=1,000,000, 2m > n, so 1 number averaging 1,000,000 sums to: 1,000,000=1,000,000

Would you like another example?  What if N is odd?  Let N=15, which has four odd divisors:

Example: N=15
When n=15, m=1, n > 2m, so 2*1 numbers averaging 15/2 sum to: 7+8=15
When n=5, m=3, 2m > n, so 5 numbers averaging 3 sum to: 1+2+...+5=15
When n=3, m=5, 2m > n, so 3 numbers averaging 5 sum to: 4+5+6=15
When n=1, m=15, 2m > n, so 1 number averaging 15 sums to: 15=15

And one more example:

Example: N = 25 33 52 7 = 151,200, which has (3+1)(2+1)(1+1) = 24 odd divisors:
When n=4725, m=32, n > 2m, so 2*32 numbers averaging 4725/2 sum to: 2331+2332+...+2394=151,200
When n=1575, m=96, n > 2m, so 2*96 numbers averaging 1575/2 sum to: 692+693+...+883=151,200
When n=945, m=160, n > 2m, so 2*160 numbers averaging 945/2 sum to: 313+314+...+632=151,200
When n=675, m=224, n > 2m, so 2*224 numbers averaging 675/2 sum to: 114+115+...+561=151,200
When n=525, m=288, 2m > n, so 525 numbers averaging 288 sum to: 26+27+...+550=151,200
When n=315, m=480, 2m > n, so 315 numbers averaging 480 sum to: 323+324+...+637=151,200
When n=225, m=672, 2m > n, so 225 numbers averaging 672 sum to: 560+561+...+784=151,200
When n=189, m=800, 2m > n, so 189 numbers averaging 800 sum to: 706+707+...+894=151,200
When n=175, m=864, 2m > n, so 175 numbers averaging 864 sum to: 777+778+...+951=151,200
When n=135, m=1120, 2m > n, so 135 numbers averaging 1120 sum to: 1053+1054+...+1187=151,200
When n=105, m=1440, 2m > n, so 105 numbers averaging 1440 sum to: 1388+1389+...+1492=151,200
When n=75, m=2016, 2m > n, so 75 numbers averaging 2016 sum to: 1979+1980+...+2053=151,200
When n=63, m=2400, 2m > n, so 63 numbers averaging 2400 sum to: 2369+2370+...+2431=151,200
When n=45, m=3360, 2m > n, so 45 numbers averaging 3360 sum to: 3338+3339+...+3382=151,200
When n=35, m=4320, 2m > n, so 35 numbers averaging 4320 sum to: 4303+4304+...+4337=151,200
When n=27, m=5600, 2m > n, so 27 numbers averaging 5600 sum to: 5587+5588+...+5613=151,200
When n=25, m=6048, 2m > n, so 25 numbers averaging 6048 sum to: 6036+6037+...+6060=151,200
When n=21, m=7200, 2m > n, so 21 numbers averaging 7200 sum to: 7190+7191+...+7210=151,200
When n=15, m=10,080, 2m > n, so 15 numbers averaging 10,080 sum to: 10,073+10,074+...+10,087=151,200
When n=9, m=16,800, 2m > n, so 9 numbers averaging 16,800 sum to: 16,796+16,797+...+16,804=151,200
When n=7, m=21,600, 2m > n, so 7 numbers averaging 21,600 sum to: 21,597+21,598+...+21,603=151,200
When n=5, m=30,240, 2m > n, so 5 numbers averaging 30,240 sum to: 30,238+30,239+...+30,242=151,200
When n=3, m=50,400, 2m > n, so 3 numbers averaging 50,400 sum to: 50,399+50,400+50,401=151,200
When n=1, m=151,200, 2m > n, so 1 number averaging 151,200 sums to: 151,200=151,200

Internet References

Related pages in this website

 

 


The webmaster and author of the Math Help site is Graeme McRae.
     [home]  [email]  [search]  [Links to Math Sites]  [Whiteboard]