Hi, rrobbo! The value for PRXX in the ratio in 4d is indeed 636. If you don't want to try all possible values of J in the other terms of the ratio (AJ+P+Q - I'm assuming that you have already solved for A, P and Q), then you could use Euclid's formula.
You can assume that 636 is equal to 2rs. It can't be r^2 - s^2, or that would make the other term equal to 2rs. Hence, both terms would be divisible by two, which contradicts the statement that the ratio is in its lowest terms, This means that rs is equal to 318.
You can factorise rs into two numbers in four different ways: 318*1, 159*2, 106*3, 53*6. However, for the first three factorisations, you will find that r^2 -s^2 has at least 5 digits, and 4d's answer has to have only four. So r is 53 and s is 6 in Euclid's formula, and you can work out what AJ+P+Q is by using the value you've calculated for r^2 - s^2, and hence get the value for J.
I suspect that no table of Pythagorean triples on the internet contains all the answers (those for 14a and 8d in particular), but the triple for 4d is on the table at
http://www.tsm-resources.com/alists/PythagTriples.txt, which gives values for triples (a, b, c) where c goes as far as 10,000.