Is carnelian an agate?
Carnelian is a colour name, agate a pattern name, both for chalcedony. Banded carnelian is both. Hardness 6.5 to 7 and SG 2.58 to 2.64 cannot tell them apart.
Only when it is banded. Carnelian is a colour name, red to orange chalcedony; agate is a pattern name, chalcedony with bands or inclusions. A banded carnelian is both, a plain one carnelian only. No test separates them: both are hardness 6.5 to 7, and carnelian’s RI 1.535 to 1.539 sits inside agate’s 1.530 to 1.543.
In short
- The colour-or-pattern test. The International Gem Society chalcedony article defines carnelian as translucent to semi-opaque, red, orange-red or brownish chalcedony, and agate by its translucency or by having a pattern to its colours. One word names a colour, the other a pattern. Ask both questions: orange-red and banded is carnelian agate; orange-red and plain is carnelian; banded but not orange-red is agate.
- The panel overlap. The International Gem Society carnelian article and the International Gem Society agate article give the same formula, SiO2, the same hardness, 6.5 to 7, and birefringence 0.004 against up to 0.004. Carnelian’s RI, 1.535 to 1.539, lies inside agate’s 1.530 to 1.543. Only one number on the carnelian panel falls outside agate’s: specific gravity 2.58 to 2.60, the bottom of its range.
- Documented failure: the translucency clause. IGS says agate may be defined by translucency, and says carnelian is semi-opaque to highly translucent. Read literally, every translucent carnelian is an agate. But the same chalcedony article says translucent single-colour material is sold as chalcedony. The translucency clause therefore cannot decide; the pattern clause can.
- Is carnelian the same as red agate? Not exactly. Red agate needs bands or a pattern; carnelian does not. IGS calls sard similar to carnelian with a brownish tone, and sardonyx onyx with white and red layers. Both carnelian and agate are chalcedony, and chalcedony is quartz: see is chalcedony quartz and chalcedony.
- Colour and enhancement, as IGS states them. IGS says iron is the source of carnelian’s colour, that it can be heat treated to darken red tones, and that any carnelian should be assumed heated unless you are told otherwise. Its agate article says dyeing is common and stable, and that many richly coloured banded agates are dyed. This page gives no test for either; dye checks are on how to tell if a crystal is dyed.
| Name | What the word names (IGS) | Colour (IGS) | Pattern needed? | Hardness | Specific gravity | 20.00 g in water | Refractive index | IMA row? |
|---|---|---|---|---|---|---|---|---|
| Carnelian | A colour: red, orange-red or brownish chalcedony | Yellow-orange to near reddish orange to orangey brown (carnelian article); red, orange (panel) | No | 6.5 to 7 | 2.58 to 2.64 | 12.25 to 12.42 g | 1.535 to 1.539 | No |
| Agate | A pattern: chalcedony told apart by pattern, translucency or inclusions | All | Yes, or translucency or inclusions | 6.5 to 7 | 2.60 to 2.64 | 12.31 to 12.42 g | 1.530 to 1.543 | No |
| Carnelian agate (banded carnelian) | Both questions answered yes | Orange-red, banded | Yes | 6.5 to 7 | Inside both ranges | 12.31 to 12.42 g | Inside both ranges | No |
| Sard | Like carnelian with a brownish tone | Carnelian colours, brownish | No | Not printed separately | — | — | — | No |
| Sardonyx | Onyx with white and red layers | White and red | Yes: straight layers | Not printed separately | — | — | — | No |
| Chalcedony | Micro- or cryptocrystalline quartz | Translucent single colours are sold under this name | No | 7 | 2.55 to 2.91 (IGS chalcedony panel) | 12.16 to 13.13 g | — | No |
| Quartz (crystal) | The IMA species, SiO2 | Colourless, white, many from inclusions (sheet) | — | 7 | 2.65 | 12.45 g | ω 1.544; ε 1.553 | Yes: A, 1967 s.p. |
Is carnelian a type of agate?
Sometimes: when it is banded or patterned. Both words sit inside chalcedony on the International Gem Society’s definitions. The carnelian article opens by calling carnelian a variety of chalcedony and gives its colour range; nothing in its definition requires bands. The agate article calls agate a variety of chalcedony whose patterns, translucency and sometimes moss-like inclusions can distinguish agates from other chalcedonies. A carnelian with bands answers to both descriptions; one without them answers to the first only.
The quotable line. Carnelian tells you the colour of a chalcedony and agate tells you its pattern, so the same stone can be one, the other or both. Neither is a mineral species: a whole-word scan of the September 2026 IMA list finds no line containing carnelian, cornelian, agate, sard, sardonyx, onyx, chalcedony or jasper. The species is quartz, and the mineral-definition test is on is quartz a mineral.
Can any test tell carnelian from agate?
No measurement in the IGS panels can. Both give SiO2, hexagonal (trigonal) micro- or cryptocrystalline, hardness 6.5 to 7, uniaxial (+), no cleavage. Carnelian’s refractive index, 1.535 to 1.539, falls inside agate’s 1.530 to 1.543. Their specific gravities, 2.58 to 2.64 and 2.60 to 2.64, read 12.25 to 12.42 g and 12.31 to 12.42 g in water for a 20.00 g piece, the same figures as on the chalcedony page. The only carnelian figure outside the agate panel is SG 2.58 to 2.60, a 0.06 g band on that weighing (12.25 to 12.31 g).
So the choice is visual. Look for bands, eyes or other patterning under a lens and in transmitted light. Orange to red colour with bands: both names. Orange to red without: carnelian. Patterned but grey, white, blue or another colour: agate. Brownish rather than orange: sard, on IGS. The separate question of whether chalcedony counts as quartz, and how to choose between the words quartz and chalcedony, is on is chalcedony quartz and is not repeated here.
Why does the translucency clause fail?
Because IGS uses translucency for both words. Its chalcedony article says agate is defined either by its translucency or by having a pattern to its colours, and in a later passage by its translucency, colour patterns or inclusions. The carnelian article says carnelian varies from semi-opaque to highly translucent. Applied literally, a highly translucent carnelian meets the agate definition with no bands at all.
The same source closes that door. The chalcedony article also says that in common practice only the translucent, single-colour types are sold as chalcedony, and that the rest are sold under variety names or as agate or jasper. A translucent, single-colour orange stone is therefore carnelian, or chalcedony, in IGS’s own usage, not agate. That is why the test on this page asks about pattern, not translucency. Two other names that turn on such definitions are on is bloodstone a jasper and is chrysoprase chalcedony.
Who is each name wrong for?
Agate is wrong for a plain, unbanded orange stone: on IGS’s pattern clause it is carnelian. Carnelian is wrong for a banded grey, white or blue agate, and for a brownish stone IGS would call sard. Both are right for a banded orange-red chalcedony, which is what a label reading carnelian agate describes.
Treatment is not a test result here. IGS says carnelian’s iron colour can be darkened by heat, even by the sun, and that you should assume heating unless told otherwise; it says agates are commonly dyed. Neither statement changes which name fits the pattern, and this page offers no way to detect either. Lore and symbolism are out of scope. More guides are under the collecting guides. We hold no catalogue and nothing here is offered for sale; the wanted list is the only commercial route.