I fancy you give me credit for being a more systematic sort of cove than I really am in the matter of limits of significance. What would actually happen would be that I should make out Pt (normal) and say to myself that would be about 50:1; pretty good but as it may not be normal we'd best not be too certain, or 100:1; even allowing that it may not be normal it seems good enough and whether one would be content with that or would require further work would depend on the importance of the conclusion and the difficulty of obtaining suitable experience.