The finder
What was taken away, and by what
Shown on purpose, not filtered out
The 4 materials here that must never go in a tank.
Three of the four are free on the ground or cheap in a builders yard in Armenia, and every one of those three sits one word or one glance away from something excellent. A catalog that quietly leaves them out is not being careful. It is withholding the part that matters.
Our own work, and it should be checked rather than trusted
Carbonate rock does not have an appetite. It has a ceiling.
Almost all the trouble comes from one mineral. Calcite is calcium carbonate and it is the main mineral in limestone7. When it dissolves in water holding carbon dioxide, one reaction produces both of the things aquarists measure. The calcium ion is general hardness. The bicarbonate ion is carbonate hardness. So a limestone stone raises GH and KH together, close to one to one in German degrees1. That is a useful fingerprint: if both climb by similar amounts, a carbonate rock is the likely cause, and if GH climbs and KH does not, something else is happening.
The hobby stops at "it raises hardness". You can do better, because calcite does not dissolve forever. It stops at saturation, and the saturation point depends almost entirely on how much CO2 is in the water. We solved the standard carbonate equilibrium system for water sitting on calcite at 25 degrees, using the usual constants12, and the table below is the result21. Activity corrections are ignored, so real numbers run a little higher, but the shape is right. The calculation checks out against the measured figure it should: USGS Water-Supply Paper 2254 gives about 20 mg/L of calcium from experiment and 16 mg/L from thermodynamic calculation1 for water in contact with air, and our solution lands at 18 mg/L21 between them.
Three consequences, and the tool above uses all three.
An uninjected tank full of limestone does not become liquid rock. It drifts towards roughly 2.7 dGH and 2.7 dKH and stops21, which is inside the range most plants and most soft water fish tolerate.
Injecting CO2 roughly quadruples what the water can hold. At 30 mg/L the target is about 9.9 dGH and 9.9 dKH21, not 2.5. The people whose Seiryu tanks keep hardening are, almost always, the people running CO2.
And water changes are the pump. Every change resets the water below saturation, so a tank on heavy weekly changes in soft water keeps pulling calcium off the rock. A tank left alone reaches equilibrium and stops.
Water at equilibrium with calcite, 25 C21
- 2.5
- dGH ceiling with no injection, open to air
- 6.8
- dGH ceiling at 10 mg/L CO2
- 8.6
- dGH ceiling at 20 mg/L CO2
- 9.9
- dGH ceiling at 30 mg/L CO2
- x4
- roughly what injection does to the ceiling
| Condition | pH at equilibrium | Calcium | Approx. GH | Approx. KH |
|---|---|---|---|---|
| Open to normal air, no injection | 8.2 | 19.4 mg/L | about 2.7 dGH | about 2.7 dKH |
| CO2 injected to 10 mg/L | 7.4 | 49 mg/L | about 6.8 dGH | about 6.8 dKH |
| CO2 injected to 20 mg/L | 7.2 | 61 mg/L | about 8.6 dGH | about 8.6 dKH |
| CO2 injected to 30 mg/L | 7.1 | 70 mg/L | about 9.9 dGH | about 9.9 dKH |
What the ceiling is not
It is not a rate.
Elephant Skin stone is limestone with much less surface area per kilogram than Seiryu22, so it works more slowly. It is heading for exactly the same place. Rate decides how long you have to notice. The ceiling decides where you end up. The tool reports the ceiling, and the cards say which materials are slow.
What boiling does
Nothing to the mineral.
Boiling kills organisms and removes loose dust. Calcite is still calcite afterwards. Nothing you can do in a kitchen changes the solubility of a rock, and the advice to boil a stone to stop it leaching is simply wrong.
Use Yerevan tap
One city, one rock, two opposite answers, and the reason is arithmetic rather than opinion.
Yerevan tap water runs 2.9 to 10.9 dGH and pH 6.01 to 7.30 across districts19, sampled from four districts and the source springs between 2002 and 2004. The soft end is Avan, Kanaker and Nork. The hard end is Davtashen. On the USGS scale that spans soft to moderately hard3. It is bicarbonate based and close to neutral because it comes through volcanic rock rather than limestone19, and it rises in summer, so test your own.
Now put that range against the ceiling. The uninjected ceiling is about 2.7 dGH21, and every district in the study is already above it So in an uninjected Yerevan tank, a limestone rock has almost nothing to do: the water is at or past saturation and there is no room left to dissolve the stone. The warnings the whole internet gives about Seiryu are, for most of this city, chemically irrelevant.
Turn the CO2 on and it inverts. The ceiling moves to about 9.9 dGH21. A scaper in Avan starting at 2.9 dGH19 can now be pulled up by seven degrees21, which will take a Caridina tank apart. A scaper in Davtashen starting at 10.9 dGH19 is still above the ceiling and still sees nothing. Same rock, same city, same advice on every forum, opposite outcomes.
That comparison is what the Yerevan tap button in the panel does. It is also the reason this hub keeps saying to buy a test kit: the difference between the two halves of the city is larger than the difference between two stones.
The rule this tool inverts on purpose
An unknown is never a yes.
The find-a-fish tool on this site holds the opposite rule, and holds it correctly. There, a species with no published hardness range stays in the results, because the databases leave that field empty for half the trade and a filter that read an empty field as a failure would delete the catalog and look confident doing it. A missing figure is not a small figure.
Hardscape is not like that, because the cost of the two mistakes is not the same. Missing a nice fish costs you a fish. Treating an unpublished stone as inert costs you the tank, slowly, over the four months it takes a Caridina colony to stop breeding. So here, unknown is a value in its own right, it is never folded into inert, and every filter that could let it through as safe holds it out instead and says why.
Four materials in this catalog carry it. Frodo, Yamaya, Manten and Koke522 are trade names with no published composition. Frodo is sold as metamorphic and reports of its acid behavior disagree. Manten is reported inert, by the people selling it. Koke's pale surface marking is consistent with carbonate, which is a hint and not a measurement. Armenian river stone is unknown for a different reason: the catchment geology is published, and which pebble you picked up is not19.
Nobody has published a petrographic analysis of any aquarium trade stone, and no thin section of a Seiryu exists in the literature. That is why every card here carries a confidence column, which is the one field the trade never prints. And unmeasured is not unmeasurable: the fourteen-day bucket test beside this paragraph settles the question for the rock in your own hands, in your own water, without waiting for anybody to publish.
The test that actually works22
- Day 0
- rock in a bucket of RO or distilled water, nothing else in it. Record GH, KH and TDS. Cover it
- Day 7
- measure again
- Day 14
- measure again
The vinegar test is a spot check, and here is what it misses.
The geological version uses dilute hydrochloric acid at 5 to 10 percent4. Household vinegar is about 5 percent acetic acid4, a much weaker acid, so the kitchen version is a weaker form of an already limited test.
Dolomite can pass it and still harden your water. Solid dolomite barely reacts to cold acid47, and dolomite is calcium magnesium carbonate, so it raises both GH and KH. Grind a corner to powder with a file or a hammer and test the dust. This matters specifically in Armenia, where dolomitic marble and Pleistocene dolomitic basalts are both documented near Yerevan1516.
Gypsum never fizzes and raises GH regardless. Gypsum is calcium sulfate1. It dissolves far more readily than calcite, it puts calcium into the water, and there is no carbonate in it to make a bubble. Any calcium sulfate or calcium chloride mineral is invisible to this test. That is not a footnote in Armenia: the Avan salt deposit is a 700 meter Miocene evaporite sequence16, and Miocene evaporites usually carry gypsum.
It is a spot test. You test the face you can reach. A calcite vein on the hidden side is not tested. Always test a freshly broken face, never a weathered outside.
A weak fizz does not tell you how fast. Rate depends on surface area, and aquascaping rock is chosen for texture, which means a great deal of surface area.
Two minutes, then two weeks. The vinegar test is the two minute version and it is worth doing on every piece, because a clear fizz is a definite answer in one direction. Silence is not a definite answer in the other. When the silence matters, do the fourteen day soak.
Where the names come from
The name is Japanese. The rock is usually Chinese. The two have been separate for years.
There is no mineral called Seiryu and no rock type called Dragon Stone. Every name the hobby uses is a marketing label attached to whatever a quarry is currently producing that looks right. There is also no customs code for aquascaping stone: it travels under the broad chapter 25 and chapter 68 headings for building and decorative stone, alongside kerbstones and worktops22. So the trade has to be read from the business to business directories where the material is actually wholesaled by the tonne.
A search of one of the two main Chinese export directories returns 30 first page products under "Seiryu Stone", and every one of them on that page comes from a single company in Hebei province10. The listed price is 350 to 600 US dollars per tonne with a one tonne minimum order10. The same directory returns 30 products under Dragon Stone and 32 under Ohko Stone10, all Chinese, all described as aquascaping rock.
Read the price again. 350 to 600 US dollars a tonne10 is 35 to 60 US cents a kilogram at the factory gate. A European aquascaping shop sells the same rock at 8 to 15 euros a kilogram. The stone is not the cost. Freight, sorting, breakage and retail margin are the cost, and that is the whole economic case for using the rock under your feet.
This does not prove that no Japanese stone exists. It does. ADA sells its own quarried layout stones under Japanese names5, and that is a real but small and expensive part of the market. And one more result is worth reporting on its own: "Frodo Stone" returns zero results on the same directory10. That is consistent with Frodo being a Western retail name invented downstream and applied to material bought under some other description. It does not tell you what the rock is. It tells you that the person who sold it to you probably does not know either, which is exactly why the tool files it as unknown.
The local advantage
Armenia is a volcanic country whose national building material is a volcanic rock, so the heavy half of a scape is lying on the ground here.
Basalt is the safe default and there is no shortage of it. It is plagioclase, pyroxene, olivine and magnetite, stable against temperature change, mechanical impact, acids, alkalis and sea salt, abundant in Shirak, Aragatsotn, Syunik and Gegharkunik and around Yerevan itself, with geological reserves put at about 125 billion cubic meters12. Inert, dense, dark, unlimited. If you want one local rock and no thinking, it is basalt.
Volcanic scoria is just as good and doubles as filter media. Armenia has about 500 Quaternary scoria cones and roughly 300 million tonnes of explored reserves13, including cones inside the city at Arinj and Avan. It is basalt andesite, a silicate with essentially no carbonate, and its internal surface area is exactly why lava rock is standard biological media everywhere in the world.
Naming trap one
The Armenian word for tuff also covers travertine, and the error is live inside Armenian Wikipedia.
Volcanic tuff is compacted volcanic ash. Armenian tuffs are liparitic, dacitic, andesitic or basaltic11, and every one of those is a silicate with no carbonate in it. That is genuinely good aquascaping stone, sold by the pallet, in pink and orange and black and red. But the article titled Տուֆ describes volcanic tuff correctly and then links to the English article "Tufa", which is a freshwater limestone11, while the English "Tuff" links to a second Armenian article, Հրաբխային տուֆ. Two articles for one rock, one of them cross linked to the wrong rock. And further down the same page it says the group տուֆ conditionally includes calcareous tuffs, that is travertine, widespread in Ararat, Shirak, Vayots Dzor and Syunik11. So a stone yard in Ararat can sell you something called տուֆ that is calcium carbonate. Pale, banded, honey or cream coloured porous stone is the warning sign, and it is only a rule of thumb. Test the stone.
Naming trap two
Say հրաբխային խարամ, never just խարամ.
On its own, խարամ means metallurgical slag, furnace waste from metal smelting, used in construction13. Armenia has copper smelter slag heaps at Alaverdi. Walk into a Yerevan builders yard and ask for խարամ and you can be handed smelter waste that looks convincingly like the excellent free volcanic rock you meant. It is one word, and it is the most expensive word in Armenian aquascaping.
Perfect and dangerous
Obsidian will not touch your water and will cut your fish.
Վանակատ is a felsic silicate glass, hardness 5, density 2500 to 2600 kg per cubic meter14, from Arteni, Hatis, Gutansar, Geghasar and Spitakasar. No carbonate, no soluble phase, no effect on GH, KH or pH. The problem is conchoidal fracture, which produces edges sharper than a surgical scalpel. It will cut your hands, a shard can slice tank silicone or an acrylic pane, and a startled Corydoras will be injured on it. Water tumbled pieces only, or blunt every edge yourself. And note the honest gap: Armenian obsidian is one of the most analysed rocks on earth14, but that work fingerprints elements locked in solid glass and no leaching or aquatic toxicity study of obsidian could be found.
The one hard rule
Never collect below Alaverdi.
Measured copper in the Akhtala river near its mouth runs 381 to 788 micrograms per liter against a catchment background of 3.0018. The Alaverdi river near its mouth runs 238 to 315 µg/L18, the Debed 4.5 km below Alaverdi runs 79 to 8618, and the Chochkan below the tailings dump 84 to 37918. Copper kills shrimp and snails at single digit microgram per liter concentrations. Also avoid the Sotk and Masrik rivers in the Sevan gold mining area, where arsenic reaches 13.16 µg/L against a background of 0.4 to 1.618. Upper Pambak and upper Debed are at background and are fine.
| Rock | Armenian | Effect on GH and KH | Verdict |
|---|---|---|---|
| Basalt | Բազալտ | None | Best default. Inert, dense, effectively unlimited |
| Volcanic scoria | Հրաբխային խարամ | None | Excellent, and doubles as filter media. Rinse hard. Never say only խարամ |
| Obsidian | Վանակատ | None | Chemically perfect, mechanically dangerous. Tumble or blunt it first |
| Pumice | Պեմզա | None | Inert. Floats until waterlogged. Best used hidden, under the substrate |
| Volcanic tuff | Հրաբխային տուֆ | Should be none | Probably fine, but soft, friable and ambiguously named. Test it |
| Calcareous tufa, travertine | Կրաքարային տուֆ | Raises both | Only for a hard water tank. Ararat, Shirak, Vayots Dzor, Syunik |
| Limestone | Կրաքար | Raises both | Only for a hard water tank. Shirak, Ararat |
| Marble | Մարմար | Raises both | Avoid. Also carries pyrite, and offcuts are resin filled |
| Sulphide mineralised rock | Toxic | Never. Kajaran and the Voghji, Alaverdi, Akhtala, Chochkan, Agarak, Sotk | |
| Metallurgical slag | Խարամ | Toxic | Never. Alaverdi smelter waste |
| Rock salt and evaporite | Քարաղ | Catastrophic | Never. Avan, north east Yerevan |
Field rules, and they are visual. Do not collect a rock if you see brassy or golden metallic specks or cubes, which are pyrite or chalcopyrite. A silvery grey flaky metallic mineral, which is molybdenite or galena. Green or blue crusts, stains or veins, which are copper and the most visible warning sign in this country. Rusty orange brown staining on a fresh break, or a sulphur smell. Orange red or lemon yellow crusts, which are realgar or orpiment, both arsenic sulphides. Or anything at all from a tailings dump, a mine spoil heap or a slag heap. Sulphide minerals oxidise in oxygenated water to sulphuric acid plus dissolved metals17, and a tank is a small, warm, well oxygenated place for that to happen in.
On the law, honestly. This hub has not found a published Armenian rule on picking up loose stone, and it will not invent one. What is on record is that protected areas cover about 387,054 ha, 13.1 percent of the country, including 232 natural monuments22, and the columnar basalt at the Garni gorge, the Symphony of Stones, is one of them22. Look at it and do not quarry it. Outside protected areas, a working quarry, and private land, picking up loose stone from a riverbed or a road cut is treated as unremarkable in practice, and that is a description of practice rather than a statement of law. A quarry will normally sell you a sack of scoria or basalt offcuts for less than the fuel it took to drive there, which removes the question entirely.
Wood
Wood does not soften your water, the tannins are gone in a day, and the white slime is supposed to be there.
It does not soften anything. Softening means removing calcium and magnesium, and wood removes neither. What wood adds is weak organic acids, which consume a little alkalinity and can pull KH down slightly. GH is untouched. And the pH only moves if your KH is already low: at 5 dKH or more a normal amount of wood tints the water and barely shifts the pH1. People who add wood to hard tap water and see no change are not doing it wrong.
The brown is not tannins. A study of red mangrove tannins found their chemical half life in natural water was less than one day8, and that they are largely removed by binding to sediment8. What keeps tinting your water is the wider pool of humic substances and dissolved organic carbon. That distinction is why a piece of wood goes on colouring the water long after the tannins should be gone.
And the brown is doing work. Researchers isolated dissolved organic carbon from the upper Rio Negro and tested it on zebrafish held at pH 4.0 in ion poor water, where it gave almost perfect protection against the ion regulation failure that low pH otherwise causes, cutting sodium and chloride losses and restoring sodium uptake that had shut down completely9. In soft acidic conditions the tint is physiology, not decoration. The older claim that tannins are antiseptic is much weaker and this page will not repeat it.
The white jelly on new wood is a good sign. Almost every new piece grows one in the first weeks. It is saprophytic growth, fungi and water moulds eating the sugars left in the outer wood, and ADA's own product notes warn buyers that Saprolegnia may appear at the initial stage6. It feeds on a limited supply and it disappears. Shrimp, otocinclus and snails eat it. Boiling before use reduces it, because boiling extracts the sugars in advance.
Sinking is a mechanical problem, not a chemical one. Wood sinks when its overall density passes that of water, which happens once the cell cavities fill. Bog preserved wood and dense tropical hardwood sink on day one. Spiderwood, manzanita, red moor and cholla do not. Soaking works and is slow, boiling is the fast version and will not sink a log, and most professionals skip both and screw the piece to a slate base. One safety note worth carrying: Rhododendron and other Ericaceae produce grayanotoxins7, concentrated in leaves, flowers and nectar rather than in old root wood, and no report of harm from aquarium spiderwood was found. It is a reason to buy wood sold for aquarium use rather than digging up a garden azalea.
For Armenia this is the hard part. Stone is free here and wood is not. The country's only specialist catalog holds one driftwood listing, a bonsai style piece at 29,000 AMD20, against 16 stone products20. Wood is the material genuinely worth importing, and the practical route is a marketplace inside the customs union rather than carrying it over a land border.
- Hem, J.D. (1985). Study and Interpretation of the Chemical Characteristics of Natural Water, 3rd edition. US Geological Survey Water-Supply Paper 2254. Public domain. Calcite solubility in contact with air, about 20 mg/L calcium from experiment and 16 mg/L from thermodynamic calculation, page 92; the hardness unit conversions; and gypsum as calcium sulfate. This is the paper our ceiling calculation is validated against. Full text PDF
- Morin, G. Understanding the General Chemistry of the Planted Aquarium. Seachem Laboratories. The equilibrium constants used in the ceiling calculation, and the statement that an alkalinity measurement does not indicate any one compound. PDF
- Hardness of Water. USGS Water Science School. The soft, moderately hard, hard and very hard classification used for Yerevan tap water. usgs.gov
- The Acid Test for Carbonate Minerals and Carbonate Rocks. Geology.com. The dilute hydrochloric acid protocol at 5 to 10 percent, the 5 percent acetic acid comparison, and the cold dolomite limitation. geology.com/minerals/acid-test.shtml
- Layout Materials: Stones. Aqua Design Amano. ADA's own descriptions of Ryuoh, Koke, Yamaya and Manten, which are the entire published record on three of the four unknown rows. adana.co.jp
- Layout Materials: Driftwood. Aqua Design Amano. The manufacturer's own Saprolegnia warning on branch wood. adana.co.jp
- Rock, mineral and plant definitions. Wikipedia entries for Limestone, Dolomite, Mudstone, Basalt, Petrified wood, Slate, Quartzite, Bog-wood, Rhizophora, Arctostaphylos, Cylindropuntia and Grayanotoxin. Used only for uncontested definitions, each of which sources its own primary literature. Limestone
- Maie, N., Pisani, O. and Jaffe, R. (2008). Mangrove tannins in aquatic ecosystems: their fate and possible influence on dissolved organic carbon and nitrogen cycling. Limnology and Oceanography 53(1): 160 to 171. The under one day half life. DOI 10.4319/lo.2008.53.1.0160
- Duarte, R.M., Smith, D.S., Val, A.L. and Wood, C.M. (2016). Dissolved organic carbon from the upper Rio Negro protects zebrafish against ionoregulatory disturbances caused by low pH exposure. Scientific Reports 6: 20377. Open access. DOI 10.1038/srep20377
- Made-in-China business to business directory, product searches. Searches for Seiryu Stone, Dragon Stone, Ohko Stone, Frodo Stone and Aquarium Driftwood, accessed 8 September 2026. The single Hebei supplier behind the first page of Seiryu listings, the 350 to 600 US dollars per tonne price with a one tonne minimum, and the zero result for Frodo Stone. Seiryu search
- Տուֆ and Հրաբխային տուֆ. Armenian Wikipedia, through the language links on both, accessed 8 September 2026. The composition of Armenian tuffs, the named quarry varieties, the statement that the group also covers calcareous travertine, and the interwiki mislabelling itself, which is the object of study rather than a citation of convenience. hy.wikipedia.org
- Բազալտ. Armenian Wikipedia, accessed 8 September 2026. Basalt mineralogy, its stability against acids and alkalis, the Armenian distribution, and the 125 billion cubic metre reserve figure. hy.wikipedia.org
- Հրաբխային խարամ and Խարամ. Armenian Wikipedia, accessed 8 September 2026. The 500 scoria cones and roughly 300 million tonnes of explored reserves, the named localities, and separately the fact that bare խարամ is metallurgical slag. hy.wikipedia.org
- Վանակատ, with Frahm, E. and colleagues (2024). Armenian Wikipedia for the physical description and the deposits; Scientific Reports 14: 9528 for the source complexes. Hardness 5, density 2500 to 2600 kg per cubic metre, and the Arteni, Hatis, Gutansar, Geghasar and Spitakasar deposits. nature.com
- Կրաքար, Տրավերտին and Մարմար. Armenian Wikipedia, accessed 8 September 2026. The three Armenian carbonate rocks, where each occurs, and the note that Armenian marble carries pyrite among its accessory minerals. hy.wikipedia.org
- Ավանի աղահանք. Armenian Wikipedia, accessed 8 September 2026. The Avan salt deposit: a 700 metre Miocene salt bearing sequence, worked since 1967, 96.84 percent NaCl, with Pleistocene dolomitic basalts in the same sequence. hy.wikipedia.org
- Քաջարանի պղինձ-մոլիբդենային հանքավայր. Armenian Wikipedia, accessed 8 September 2026. The Kajaran copper and molybdenum deposit in the upper Voghji basin, its ore minerals, and the sulphide belts at Teghut, Agarak, Shamlugh and Sotk. hy.wikipedia.org
- Gevorgyan, G.A. and colleagues (2016). Environmental Risk Assessment of Heavy Metal Pollution in Armenian River Ecosystems. Polish Journal of Environmental Studies 25(6): 2387 to 2399. Every copper, zinc, cadmium, arsenic and nickel figure on this page, including the 3.00 microgram per litre copper background and the 788 microgram per litre Akhtala maximum. pjoes.com PDF
- Shahinyan, H.V. (2005). On the hydrogeochemical characteristics of some drinking waters supplying the city of Yerevan. Proceedings of NAS RA, Earth Sciences LVIII(1): 41 to 45. Sampling 2002 to 2004, four districts and the source springs. GH 2.9 to 10.9 and pH 6.01 to 7.30, and the Hrazdan catchment geology, including the carbonate outcrops in the Argichi, Vedi and Hrazdan basins. arar.sci.am PDF
- Gazanik category feeds and product pages. The shop's own published catalogue and its live category pages, 8 and 9 September 2026. The 16 stones, the single driftwood listing at 29,000 AMD, the Dragon Stone price around 2,400 AMD and the 2,200 to 3,000 AMD per kilogram band. Individual prices are indicative. gazanik.am
- Our own arithmetic. Not a published source. The carbonate ceiling table, and every ceiling figure derived from it anywhere on this page or inside the tool. Method: the standard calcite carbonate equilibrium system solved at 25 degrees C for a fixed dissolved CO2, charge balance 2[Ca2+] = [HCO3-], using log K1 -6.35, log K2 -10.33 and log Ksp(calcite) -8.48, with activity corrections ignored so real values run slightly higher. Constants from sources 1 and 2. Validated against the calcite solubility figures on page 92 of source 1. The same calculation drives the live figure the tool prints on each card. calculated, repeatable, unverified by a third party
- the stone table and the wood table with their per row confidence ratings; Armenian stone, and Yerevan tap water; ; hardscape trade; concrete and the Armenian stone option. Each of those files carries its own numbered source list to the primary material, and the protected area figures come from Armenia's 5th National Report to the Convention on Biological Diversity, cited there. The hub's sources page
Continue
Where to go next.
The whole hardscape table
Every material with every field showing, sortable, and each row carrying its own source ids.
Materials and water
The carbonate ceiling in full, which is what decides whether a rock matters at all.
Armenia is standing on its hardscape
The Armenian side of it. A volcanic country whose national building stone is aquascaping stone, quarry by quarry and colour by colour.
Stone, hard water and the filter
Why the wrong rock is mostly not a disaster, with the arithmetic rather than the folklore.
Chemistry
The carbonate system solved, which is the chemistry every row in the table depends on.