What goes wrong

It leaked, or the sample was not the sample

On this page11

The sample bottle had headspace The container leached into the sample Which is the argument for glass on anything measured The science fair sample went through a pressure cycle The glass jar met a hard floor The crystal grew into the neck Or it never grew, because the jar was sealed The reagent degraded in the light The cap was cross-threaded by a student Test it the way a class will use it References

A container can change the thing it holds. For a sample that is the whole failure.

The sample bottle had headspace

A water pH drifts as carbon dioxide exchanges with the air above it. A bottle filled to the brim and capped gives a reading closer to the source.

The container leached into the sample

Or adsorbed something out of it. For a measurement this is the worst failure mode, because the result looks fine and is wrong.

Which is the argument for glass on anything measured

130 glass containers. Inert, and the default where the number matters more than the convenience.

The science fair sample went through a pressure cycle

A warm car, a cold hall, and a cap tightened by hand. Leakproof means leakproof through that cycle, not on the bench where it was filled.

The glass jar met a hard floor

In a room of thirty. For a specimen jar that is an exposure event rather than a breakage, which is why storage height and trays are part of the specification.

The crystal grew into the neck

And the only way out was breaking the jar. A wide straight sided jar is the right growth chamber and a bottle is not.

Or it never grew, because the jar was sealed

Evaporation is the mechanism. A proper seal on a crystal jar stops the experiment, which is the one place in this catalog where a worse closure is the right answer.

The reagent degraded in the light

Indicators and biological stains both. 54 amber containers here, and a closed cupboard costs nothing.

The cap was cross-threaded by a student

And leaked in the cupboard over a week. A closure a fourteen year old can close properly is a safety specification, not a convenience.

Test it the way a class will use it

Filled, capped by somebody in a hurry, carried, put away, opened a week later. Every failure on this page appears in that test and none of them appear on the day you buy it.

References

The Davis Pack Supply catalog, measured 18 October 2026. The classroom and home lab shelf is 987 rows: 414 containers and 573 closures. 5 mL to 946 mL with a median of 120; 71 containers at 30 mL or under, 211 at 120 or under, 335 at a pint or under. 154 HDPE, 130 glass, 117 PET, 13 PP. 259 bottles, 150 jars, 3 jugs, 2 tubes. 109 white, 105 clear, 57 natural, 54 amber. Closures: 373 continuous thread, 44 child resistant, 32 dropper assemblies, 27 dome lids, 27 spouted, 19 flip top, 14 phenolic, 10 liners and sealing discs, 9 specialty, 7 buttress, 3 dropper tips and caps, 3 euro droppers.

The archetype behind this shelf is ind-lab: 10 to 1,000 mL, narrow mouth, glass and HDPE, cold fill, thin, aqueous, with no dispensing requirement and child resistance recorded as not required. Its stated reason is "reagent or sample. Autoclavability and chemical resistance." Our catalog does not record autoclavability for any product, and of the two materials named only glass is generally autoclavable.

Written from 29 CFR 1910.1048 and from the catalog as it stood on 18 October 2026. No autoclave cycle is quoted and no container is described as autoclavable, because our data does not record it. School district and state rules on classroom chemicals exist, vary, and are not summarized here.