The percentage composition of peridot, though the ferrous oxide may be more or less replaced by magnesia, is approximately:
| Silica | 41 | Ferrous oxide | 9 |
| Magnesia | 50 | ||
Manganese, nickel, and lime have been found in small quantities in some olivines.
The term olivine is now wrongly applied by dealers in precious stones and by jewellers to the green garnets or demantoid from Bobrovka; these, for a time at least, were called, with equal incorrectness, Uralian emeralds. A considerable quantity of pale green cut gemstones has been sold in London as peridot, but on examination proved not to be that variety of chrysolite. They were softer than peridot, and, though more glittering in lustre, were of poorer colour. They were not dichroic and showed no sign of crystalline structure. On analysis they were found to contain much more silica and much less magnesia than peridot, while alumina and soda were present in distinct quantities; if the material had not been known to be a natural product, these specimens would have been called green glass. Their composition varies widely, and so does their density, the latter ranging from 2·36 to 2·63. The mineral which occurs in several Bohemian and Moravian localities is known as water-chrysolite, psendo chrysolite, moldavite and bouteillenstein. Various opinions are held as to its origin, some mineralogists even going so far as to state it to be an artificial glass from the site of early glass factories. It has also been conjectured that it may be of meteoric origin. Some glass has undoubtedly been sold as moldavite, but the genuine material differs in intimate structure, in fusing point and in chemical composition from any kind of glass. It is not identical with obsidian, while its occasional occurrence as rolled pebbles in the gem-gravels of Ceylon strengthens the view that there is a moldavite which is a natural product.
