dtweah wrote:In the calculation of a/b the following remainder sequence was observed: 3264513.... On the basis of this the following propositions were made:
I. a +b > 12
II 9<a+b<18
III b>a
It took me several minutes just to figure out what you meant by this question, since as written it doesn't make sense; when you divide a by b, of course you will always get the same remainder, and not a sequence of different remainders. As best I can tell, you don't mean "in the calculation of a/b", but rather "in the calculation of a
^n/b, for n = 1, 2, 3, ...". Then if you let a = 3, for example, when you divide 3^1 by 7, the remainder is 3; if you divide 3^2 by 7 the remainder is 2; if you divide 3^3 by 7, the remainder is 6, and so on. That produces the sequence in the question: 3, 2, 6, 4, 5, 1, 3, ... ad infinitum. Of course, a need not be 3 here; it could be any number with a remainder of 3 when divided by 7: a could be 10, 17, 24, etc.
That said, this:
dtweah wrote:
1/7 has remainder 1326451
doesn't accord with the above interpretation at all. The powers of 1 will produce remainders of 1 every time, and the sequence will be 1, 1, 1, 1, ...
dtweah wrote:
There are certain primes whose remainders exhibit a periodicity that correlates with the prime itself. 7 is one of them.
There aren't just 'certain' primes that exhibit this periodicity; this is true for all primes. If you take any prime p, you can always find a positive integer x less than p for which x^1, x^2, x^3, ... x^(p-1) produce all possible nonzero remainders when you divide by p. When p = 7, both x = 3 and x = 5 produce all six possible nonzero remainders when raised to powers 1 through 6, for example. When p=11, x=2 will produce all ten nonzero remainders when raised to the powers 1 through 10. One can find similar examples for every prime. This is a famous result of Gauss, and is a fundamental theorem in Algebraic Number Theory.
Still, it will never be tested on the GMAT, since no GMAT test taker would ever be expected to know it. It is certainly an interesting property of numbers, I'll agree.