Graduated cylinders and their tolerance

The symptom is a standard that will not repeat, or two careful people who cannot agree on the same cylinder. The cylinder is often innocent of mystery and guilty of being used past its tolerance. This troubleshooting note is for that disagreement. Cylinders sit between beakers and volumetric flasks in choosing laboratory glassware: useful for an approximate transfer, unfit for declaring a primary concentration.

What the symptom usually means

Start with how the volume was read, before you blame the glassblower. A cylinder has a wide enough bore that parallax is obvious once you look for it. Eye above the liquid, and the meniscus appears to sit on a lower line than it does. Eye below, and you over-read. The cylinder tilted in a hand changes the height of the liquid against the scale. The scale was printed for a vertical wall.

Next, ask whether this volume should have been a cylinder at all. Preparing "exactly 100 millilitres" of a stock in a cylinder imports the cylinder's tolerance into every later dilution. That tolerance is wider than a one-mark volumetric flask's. The standard that defines the cylinder says how wide. NIST's Office of Weights and Measures is the public route into how laboratories treat calibration. If the method needs flask tightness, the fix is a flask, not a steadier hand.

Likely causes and the next check

Parallax is the first check. Put the cylinder on a flat bench. Lower your eye, or raise the cylinder on a stand, until you look along the line. For aqueous solutions read the bottom of the meniscus. The top of the curve is the wrong edge for that convention and can differ by a noticeable fraction of a small cylinder.

The second check is to contain versus to deliver. Some cylinders are calibrated to contain (In). Some are calibrated to deliver (Ex), which assumes a drain. If you fill an In cylinder and pour it out, the receiving vessel does not get the full marked volume, because a film remains. If you use an Ex cylinder as if it contained its reading after a sloppy pour, you have double-counted the film. Read the letters on your cylinder.

The third check is dirt and etch. Detergent and grease distort the meniscus until it is not a clean curve. Alkali etch fades the printed lines so two readers interpolate differently. A chipped base makes the cylinder stand slightly off vertical. Any of those is a reason to wash properly or to retire the piece, not to average two guesses.

The fourth check is temperature. Cylinder calibrations also assume a reference temperature. A hot liquid from a drying step is not that liquid. Let it cool when the volume matters. For rough media preparation the effect can be smaller than other errors. For a standard, you should not be in the cylinder.

The fifth check is the wrong size. Reading 8 millilitres in a 500 millilitre cylinder uses the part of the scale where the subdivisions are coarsest relative to the volume you want. Use a cylinder whose top mark is near the volume you need, and accept that "near" is still approximate.

After the reading is honest, the tolerance remains

A clean, level, eye-height reading still carries the cylinder's class tolerance, which is built wider than a flask's because of the geometry. Recording 47.5 when the subdivisions are 1 millilitre pretends the glass resolved a quarter. Interpolate only as far as the scale honestly allows, and write the method as approximate if that is what you did.

Plastic cylinders remove breakage and add a compatibility question. Polypropylene is a common aqueous choice. It can be harder to see a meniscus in, and it is not a solvent-proof default. Glass cylinders for solvent work should be a stated glass, with borosilicate preferred when the piece may be warmed. ASTM E438 is the specification language for those glasses. Soda-lime cylinders crack more readily from thermal shock and should not be promoted to a hot plate because they once measured water.

If the cylinder is the right tool and the volume still disagrees with a balance, check density and temperature before you send the cylinder for repair. A 100 millilitre water portion weighed on a balance is a useful sanity check when the method is allowed to use one. It is not a home-made recalibration you scribble onto the glass.

Subdivisions, class marks, and a check that is not a recalibration

A class printed on a cylinder, when one is printed, is the tighter or looser band in the cylinder's own standard. It is still a wider promise than a one-mark flask holding a single volume. Read the class off the glass. If no class is printed, do not invent one because the moulding looks sharp. Record the cylinder as an approximate transfer and move any concentration you must declare into a volumetric flask.

Subdivisions set the last honest digit. A cylinder marked every millilitre does not resolve a tenth of a millilitre. A digit past the subdivision turns a rough pour into a false standard. If two people must agree more tightly than those lines allow, a perfect meniscus reading still leaves you in the wrong vessel.

A balance can sanity-check a water portion when the method allows a mass and a temperature. Use that check to decide whether the reading technique is at fault. Do not scratch a new line on the glass and call the cylinder recalibrated. Recalibration is a recorded metrology activity. If a careful reading and a sane mass still disagree with a certificate you actually hold, retire that cylinder from any job that cited the certificate.

Safety

Cylinders full of solvent tip easily because they are tall. Use a carrier, and keep the volume in a tray the chemical requires. A broken cylinder is sharps plus whatever it held. Biological liquids make it a biosafety object under your rules, with the WHO biosafety manual and the CDC BMBL as background reading. This is not a diagnostic measuring procedure.

What to specify next time

Nominal capacity near the volume you actually dispense, glass or a named resin, to contain or to deliver, and the class if a certificate matters. Say plainly if the true standard is a volumetric flask, so the cylinder is not asked to impersonate one. The laboratory glassware catalogue and the volumetric glassware catalogue are the places to separate those lines, and the quote request is where the count of pieces goes.