The composition of valuable ancient metal objects, especially those composed of alloys of gold or silver, often cannot be estimated by chemical methods because samples cannot be taken. The utility and reliability of specific gravity as an index of the composition of such objects are critically discussed with numerous examples. It is generally useful for qualitative purposes, especially for distinguishing objects composed entirely of base metals. For estimating gold in gold objects it is reliable and sufficiently accurate when the proportion of good is high, but increasingly less so as the proportion of gold becomes lower. It is reliable for estimating silver in silver objects only when the proportion of silver is very high. No quantitative significance can be attached to the specific gravity of ancient objects composed of alloys of base metals. In spite of limitations, specific gravity is a valuable index of the chemical composition of certain kinds of ancient metal objects.