Frank, you were almost certainly playing with a stacked deck. Someone put #24 in first, then #23, etc.
If everything were completely random, then the chances that you'd pull the correct part on the first draw was 1 in 24 -- 1/24 in Math. Then the chances that you'd pull the next correct part on the second draw was 1 in 23 ('cause now there's 23 parts left in the box) or 1 in 23. The probability SO FAR of getting it right is 1 in (24)(23). Keep doing that, and when you're done the probability is 1 in (24)(23)(22) ... (2)(1) (always remember to multiply by one at the end -- it gives you a feeling of confidence that's always unwarrented, but on some days can be very comforting).
The number (n)(n-1)(n-2)...(2)(1) is special -- it's called "n factorial", and noted as "n!" I'm not sure why they chose '!' -- perhaps because n! grows so astonishingly fast.
At any rate, 24! is a bit more than 6.2 x 1023, or in English 620 sextillion (I had to look up "sextillion". It's big. Almost as big as a bazzilion, which is a system's engineering term for "bigger than you can imagine, but not quite infinite"). To put that into context, if you started doing that experiment today with truly randomized boxes of stuff, then on average you could expect the sun to have gone out and possibly the entire universe to have gone dark before you hit the jackpot and got a perfect draw (assuming, of course, than you hadn't gotten bored and quit yet).