Each percentage carries the counting error from Van der Plas and Tobi (1965): σ = √(p(1−p)/n). More points narrow it, and nothing else does.
This is why 300 to 500 is the convention: it buys about ±5% on a major mineral. Chasing ±1% costs thousands of points, and the section is rarely that representative of the rock anyway.
Enter the mass retained on each sieve, coarsest first. Everything below is Folk and Ward (1957), the graphic measures a sedimentologist expects to see.
Dry the sample, weigh it. Soak it until saturated, blot the surface, weigh again. Then weigh it hanging in water. Three masses give bulk density, grain density and porosity.
The temperature matters: water is 998.20 kg/m³ at 20 °C but 997.05 at 25 °C, and ignoring that puts about 0.1% into your density. Kell's equation is used, not a constant.
Acid into water, never water into acid. The heat of dilution can boil the surface and throw acid back at you.
meq/L is what the water chemistry module needs for a charge balance.
Hydrofluoric acid
HF does not hurt when it burns you. It soaks through skin without pain, binds the calcium in your blood, and can stop a heart hours later. A concentrated burn over an area the size of your palm has killed people.
▪ Calcium gluconate gel must be in the room and in date before the bottle is
opened. Not in a cupboard down the corridor.
▪ Fume hood, face shield, neoprene or butyl gloves — nitrile alone is not enough.
▪ Never work with it alone, and tell somebody you are starting.
▪ Any contact at all: flood with water, apply the gel, go to hospital and say the words
"hydrofluoric acid", even if it does not hurt. Especially if it does not hurt.
HF is used to etch a slide before staining for feldspar. Many labs no longer allow it and use a cold sodium cobaltinitrite method instead. If you have the choice, take it.
Alizarin red S — calcite from dolomite
0.2 g alizarin red S in 100 mL of 0.2% cold hydrochloric acid. Immerse the polished surface or uncovered section 10–15 seconds, rinse gently in distilled water, air dry.
Calcite stains deep pink to red. Dolomite stays unstained. That single difference has settled more carbonate petrography than any instrument.
Potassium ferricyanide — the iron-bearing ones
2 g potassium ferricyanide in 100 mL of 0.2% HCl. Usually combined with alizarin red S in one bath, mixed 3 parts alizarin to 2 parts ferricyanide.
Ferroan calcite turns mauve to purple; ferroan dolomite turns turquoise blue. Non-ferroan dolomite still refuses to stain, which is how you tell all four apart on one slide.
Potassium ferricyanide releases hydrogen cyanide if it meets strong acid. Keep the solution dilute, work in a hood, and never store it acidified.
Sodium cobaltinitrite — potassium feldspar
Etch the surface in HF vapour for 10–15 seconds, rinse, then immerse in saturated sodium cobaltinitrite for about a minute and rinse.
K-feldspar turns bright yellow. Plagioclase and quartz stay clear, which is what makes a point count of a granite possible with a hand lens instead of a microscope.
This is the step that needs HF. Read the panel above.
Amaranth or barium chloride — plagioclase
After the cobaltinitrite, rinse in barium chloride then stain with amaranth (rhodizonate) solution.
Plagioclase takes a pink to red stain, so a stained slab shows yellow K-feldspar, red plagioclase and colourless quartz — the three corners of QAPF, visible by eye.
Nital — the Widmanstätten pattern
2–5% nitric acid in ethanol. Polish the cut face of an iron meteorite to a mirror, swab with nital for 10–60 seconds, stop in ethanol, dry immediately and warm.
The interlocking bands appear because kamacite and taenite etch at different rates. They take roughly a million years per degree of cooling to grow, so they cannot be faked — see the meteorite tests.
Nitric acid in ethanol is an oxidiser with a fuel. Make it fresh, keep it small, never seal it in a bottle, and never store it.
▪ Perchloric acid — forms explosive perchlorates with organic matter. It
needs its own washdown hood, not the general one.
▪ Sodium polytungstate — heavy liquid for mineral separation, far safer than the bromoform
and methylene iodide it replaced. If your lab still has those two, they are toxic and should go.
▪ Hydrogen peroxide 30% — used to remove organics before sieving; burns skin white on
contact.
▪ Anything unlabelled — dispose of it. A geology lab accumulates bottles for decades, and a
bottle whose label has faded is not a reagent, it is a hazard.
Entries are timestamped and appended, never edited — that is what makes a notebook evidence rather than a document. Corrections go in as a later entry.
Kept on this phone only. Export it somewhere safe now and then.