ramdomness of two independent samples

I´ve two small and independent samples and i need to confirm that they are random (one of the assumptions on the mann-whitney test), the only test i see available is the wald-wolfowitz test, but i´m not convinced, because it tests a lot of things...
Can someone give me a valuable argument!
I really feel in the gold age of my ignorance!:shakehead
Thanks a lot for your help! It confirms what I've done. So, i applied a run test for each sample - as suggested in Kanji (1993: 108)
Best regards,