How can counting arrangements explain a tendency to mix?

When gas is released into an empty part of a container, it spreads into the available space. Individual molecules do not consult a plan telling them to distribute evenly. Their motion and collisions carry them through many arrangements. Counting those arrangements reveals why a broadly spread-out state can be overwhelmingly more likely than a state crowded into one region.

Begin with just four particles, labeled A, B, C, and D for bookkeeping. Divide the available space into two equal regions and record only whether each particle is left or right. Each label has two possibilities, so there are 2 × 2 × 2 × 2 = 16 assignments. Assume the particles' side choices are independent and each side is equally probable. Under those assumptions, each of the sixteen assignments has probability 1/16. The labels are not four chemical species, and the record leaves out positions within each region and particle energies.

A broad description, or macrostate, specifies only how many particles occupy each side. The macrostate with all four on the left has one compatible assignment: ABCD on the left and none on the right. Two on each side has six: AB, AC, AD, BC, BD, or CD can form the left-hand pair, with the other two labels on the right. Each detailed assignment is equally likely, but the two-and-two macrostate is six times as likely as all four on one specified side.

Counting all five possible left-hand totals gives 1, 4, 6, 4, and 1 assignments. They add to sixteen, so no possibility has disappeared. The two-and-two outcome has probability 6/16, or 37.5 percent. It is the most likely single total, but it is not more likely than all other totals combined. With only four particles, uneven divisions remain common. Statistical reasoning allows fluctuations; it does not command every snapshot to look perfectly balanced.

As the number grows, a sharply crowded state becomes much harder to encounter. With one hundred independent particles under the same assumptions, all on the left has probability 1/2¹⁰⁰, about 7.9 × 10⁻³¹. Exactly fifty on each side still is not certain: it has probability about 8 percent. The important concentration is across many totals near half, not just one exact split. An ordinary gas contains vastly more particles, making large relative imbalances far less typical at equilibrium.

Multiplicity, written W, means the number of compatible arrangements. The relation S = k ln W connects entropy with that number, where k is Boltzmann's constant and ln is the natural logarithm. For the restricted four-label count, the two-and-two macrostate has S/k = ln 6, about 1.79; a specified all-on-one-side state has ln 1 = 0. These are comparisons within this deliberately limited description, not absolute entropies of actual gases.

The count alone does not tell us how fast spreading occurs. A physical explanation also needs accessible motion, energy constraints, and time. Nor does entropy simply measure how untidy an image looks. Its deeper lesson is that many microscopic possibilities can produce one familiar large-scale result. A room can appear steady while an enormous number of molecular arrangements continually replace one another.