An inch is not an inch
Layer plans are written in inches because inches are what you can see through the glass. Materials are bought by volume. In a straight-sided jar or a rectangular tank the conversion is one multiplication and nobody gets it wrong. In anything curved, the same inch is worth a different amount depending on where it sits.
A 10 inch globe makes the point better than any explanation. The bottom inch of it holds 14.7 cubic inches. An inch taken across the widest part holds 78.3 cubic inches, 5.3 times as much. So a 1 inch drainage layer costs almost nothing, and the 2.5 inches of substrate above it — sitting where the sphere is at its fattest — costs 143.2 cubic inches, which is 86 percent of everything in the container.
The arithmetic behind that is the spherical cap: the volume filled to a depth d in a sphere of radius R is pi times d squared times three R minus d, all over three. The page applies it to each layer boundary and takes the difference, so each layer gets its own true volume rather than a share of the total.
What the curve saves
Take the default layers — 1 inch, a quarter inch and 2.5 inches, 3.75 inches in total — and imagine the container were a straight-sided cylinder 10 inches across for its whole height. That would take 294.5 cubic inches. The globe takes 165.7. The curve saves 128.8 cubic inches, about 44 percent, which is the difference between one bag and two.
People discover this in the other direction and more expensively: they estimate a globe as if it were a jar, buy accordingly, and have most of a bag left over. Or they estimate a wide bowl by its rim diameter, which is the widest point, and overbuy by even more.
Headroom is the thing you are actually building
The layers are infrastructure. What you are making is the space above them, and in a curved container that space is smaller than the vertical measurement suggests, because the walls are already closing in above the halfway line. The defaults leave 6.25 inches of vertical room in the globe, and it is still widening for only the first 1.25 inches of that, up to the equator. Everything above the equator narrows toward the opening.
The page reports the headroom as a volume as well as a height for exactly this reason. 357.9 cubic inches of space above 165.7 cubic inches of layers is a different sort of statement from "6.25 inches of room", and it is the one that tells you whether something will actually fit.
The opening is a separate constraint
A 4 inch opening on a 10 inch globe means everything that goes in has to pass through a circle of 12.6 square inches, tools included. The material itself is not the problem — 165.7 cubic inches is about 11.5 cups, and a funnel or a rolled sheet of paper handles that. The problem is anything rigid, and anything that has to come back out later.
The page reports the opening area and the ratio of the widest inside point to it, which is the number that tells you what sort of build this is going to be.
Weights, and what the page will not say
Density is the input people are most tempted to skip and the one most worth measuring. Fill a measuring cup, weigh it on a kitchen scale, and multiply grams per cup by 0.264 for pounds per cubic foot. The defaults are placeholders. Materials sold under the same name vary by a lot depending on how wet and how compacted they arrive, and the difference between 80 and 110 pounds per cubic foot for gravel is a third of the weight of the bottom layer.
What layers to use, how deep, what goes in the container and how it is watered or lit are not questions this page answers. It converts depths into volumes and weights for a shape you describe. A glass container full of damp material is heavy and breaks into sharp pieces; damp enclosed material grows organisms that matter more to anyone whose immune system is compromised; and many common houseplants are toxic to cats, dogs and small children, which is a question for a vet, a doctor or poison control rather than for arithmetic.
Questions people ask
Why does a globe need different maths from a jar?
Because the cross-section changes with height. In a jar or a tank, every horizontal slice is the same size, so volume is area times depth and a layer plan in inches converts by multiplication. In a sphere the slice starts at nothing, grows to its widest at the halfway point, and shrinks again, so the volume filled to a given depth follows the spherical cap formula rather than a straight line. On a 10 inch globe the practical consequence is that the bottom inch holds about a fifth of what an inch across the middle holds, and any plan that treats them as equal will be wrong in both directions at once.
How do I measure the inside diameter of a globe with a small opening?
Measure the outside across the widest point with a tape or a pair of rulers held parallel, then subtract twice the glass thickness. Glass on a decorative globe is commonly somewhere between an eighth and a quarter inch, and you can usually see the edge at the opening well enough to judge it. Being out by an eighth of an inch on a 10 inch globe moves the layer volumes by under two percent and the brim-full figure by about four, which is inside the error of most of the other inputs anyway.
What density should I use for my materials?
Yours, measured. Fill a measuring cup with the material as it will go in, weigh it on a kitchen scale in grams, and multiply grams per cup by 0.264 to get pounds per cubic foot. The defaults on the page are placeholders for washed gravel and a bagged mix and should not be treated as facts: the same nominal material varies with grain size, moisture and how compacted it was in the bag. If you only want volumes and not weights, clear the density fields and the weight section is skipped.
The calculator cut my layers off. What happened?
The depths you entered add up to more than the inside height of the container, so the page truncated them at the top and told you. The figures shown then describe what actually fits, not what you asked for. This happens most often with globes, because the diameter is both the width and the maximum fill depth, and a layer plan sized for a tank of the same nominal size will not fit. Reduce a depth or use a larger container.
Does this work for an open bowl rather than a closed globe?
For the part that is a sphere, yes. A bowl that is a hemisphere is a sphere filled to no more than half its diameter, and the spherical cap arithmetic handles that exactly — enter the bowl diameter as the globe diameter and keep the layer total under half of it. A bowl that flares outward at the rim rather than curving back in is a different shape and the page will underestimate the volume near the top, though at typical layer depths that region is above everything you are filling anyway.