Kinetic Theory of Ideal Gases

Molecular collisions and gas pressure

Kinetic theory models a dilute gas as a large collection of molecules in continual random motion. Between collisions, each molecule moves approximately in a straight line at constant velocity. Collisions with the container walls reverse the velocity component normal to a wall and transfer momentum to that wall. Gas pressure is the macroscopic effect of the enormous number of such momentum transfers per unit area and per unit time. A molecule moving parallel to a wall does not transfer momentum to that wall until a collision changes its direction.

An elastic wall collision changes the momentum of a molecule of mass $m$ approaching a wall normal to the $x$ axis with velocity component $v_{x}$ from $m v_{x}$ to $- m v_{x}$ . The molecule therefore has momentum change $- 2 m v_{x}$ , and the wall receives the opposite impulse $+ 2 m v_{x}$ . In a cubical container of side $L$ , the same molecule reaches the same wall at intervals $2 L /∣ v_{x} ∣$ . Its average force contribution is consequently proportional to $m v_{x}^{2} / L$ . The square is essential: molecules arriving from opposite directions both increase the average pressure.

Summing the contributions of all molecules gives

p V = \frac{1}{3} N m \overline{v^{2}} = \frac{2}{3} N \overline{K}_{tr},

where $N$ is the number of molecules, $V$ is volume, and the bar denotes an average over molecules. The factor one third follows from isotropy. In a gas with no preferred direction, the averages of $v_{x}^{2}$ , $v_{y}^{2}$ , and $v_{z}^{2}$ are equal, so each contributes one third of the average squared speed. A gas streaming through a tube or a strongly directional molecular beam does not satisfy this isotropic assumption without further analysis.

Gas pressure is the accumulated recoil of normal molecular impacts on the wall. A faster incident molecule, or a higher collision rate, raises the average force on the shaded wall.

Temperature and translational kinetic energy

The ideal-gas law gives $p V = N k T$ , where $k$ is Boltzmann's constant. Comparing this relation with the kinetic-theory pressure equation gives

\overline{K}_{tr} = \frac{3}{2} k T .

Temperature on the kelvin scale therefore fixes the average translational kinetic energy per molecule, not the kinetic energy of every individual molecule. Two gas samples at the same temperature have the same average translational kinetic energy per molecule even when their molecules have different masses. Their speed scales are different: lighter molecules must move faster to carry the same average kinetic energy. Temperature is not a measure of the total kinetic energy of a sample, because the total also depends on the number of molecules.

The root-mean-square speed is defined by $v_{rms} = \overline{v^{2}} = 3 k T / m$ . It is the relevant speed scale because pressure depends on the average of $v^{2}$ , but it is not the average speed and it is not the most probable speed. For one mole of gas with molar mass $M$ , the equivalent form is $v_{rms} = 3 R T / M$ . At the same temperature, helium has a greater rms speed than oxygen because its molar mass is smaller.

The distribution of molecular speeds

An ideal gas at equilibrium does not contain molecules all travelling at one speed. The Maxwell--Boltzmann speed distribution places a fraction $f (v) d v$ of the molecules in the narrow speed interval from $v$ to $v + d v$ . Its functional form is

f (v) = 4 π (\frac{m}{2 π k T})^{3/2} v^{2} exp (- \frac{m v ^{2}}{2 k T}) .

The distribution begins at zero speed because the number of distinct velocity directions associated with a very small speed is small. It rises to a peak, then falls with a long high-speed tail. The most probable speed is the peak location, $v_{mp} = 2 k T / m$ . The average speed and rms speed lie farther to the right because rare fast molecules contribute strongly to averages, especially to the squared-speed average. The ordering is $v_{mp} < v_{av} < v_{rms}$ .

Increasing temperature shifts the peak to higher speed, lowers its height, and broadens the curve. The total area remains one because every molecule belongs to some speed interval. At a fixed temperature, increasing molecular mass shifts the distribution toward lower speeds. These curve changes explain why a single molecular speed cannot represent all collision rates, diffusion processes, or escape probabilities. The rms speed is appropriate for pressure calculations; the full distribution is needed when the high-speed tail matters.

A Maxwell–Boltzmann speed distribution has a peak and a long high-speed tail. Higher temperature shifts the peak rightward and broadens the curve while the area under each stays one.

Kinetic theory describes dilute gases accurately when intermolecular forces and molecular volume are small corrections. Near condensation, at high density, or in strongly non-equilibrium flows, the ideal-gas assumptions may fail. The pressure relation and speed distribution then require a model that includes molecular interactions or a measured equation of state. Within its stated range, the theory connects pressure, temperature, and molecular motion quantitatively.

The pressure formula contains an average over a very large molecular population. An individual molecule exerts a sequence of short impulses separated by free flights, so its force on a wall is not constant. The measured pressure becomes nearly steady because a macroscopic wall receives collisions from many molecules at many locations and times. Reducing the number density lowers the collision rate at the wall and lowers pressure at fixed temperature. Reducing the volume at fixed number of molecules raises the collision rate because the travel distance between opposite walls is shorter. These two microscopic effects are contained together in the inverse-volume dependence of the ideal-gas law.

Molecule--molecule collisions maintain the equilibrium distribution by continually redistributing velocities. The mean free path is the average distance travelled between such collisions. For molecules of effective diameter $d$ and number density $n$ , the hard-sphere estimate is $λ = 1/ (2 π d^{2} n)$ . The factor $2$ accounts for the fact that the other molecules move as well. At ordinary gas density, the mean free path is much larger than a molecular diameter, which supports the idealization of long free flights interrupted by brief collisions. At higher density, molecular separation and collision duration become less negligible, and the simple ideal-gas picture loses accuracy.

The vertical axis on a speed-distribution graph is a fraction per unit speed, not a fraction at one exact speed. A narrow interval has fraction approximately $f (v) Δ v$ , represented by an area under the curve. Consequently, curve height alone does not compare populations unless the speed intervals have equal width. The broad high-temperature curve can have a lower peak than the low-temperature curve while containing the same total fraction of molecules. Integrating over any chosen speed range gives the fraction relevant to a detector threshold, an escape-speed estimate, or a collision-rate calculation.

Pressure from momentum transfer

The pressure derivation can be written as a sequence of momentum and timing statements. One molecule occupies a cube of side $L$ ; take the wall normal to the $x$ direction. Immediately before an elastic collision its momentum component is $m v_{x}$ ; immediately after, it is $- m v_{x}$ . The molecule's change is $- 2 m v_{x}$ , so the wall receives impulse $2 m v_{x}$ in the positive direction. The approaching molecule has the velocity sign appropriate to the selected wall, and the transferred momentum always pushes outward on that wall.

The molecule must travel to the opposite wall and back before striking the same wall again. Its return time is $2 L /∣ v_{x} ∣$ . Dividing the impulse magnitude by this time gives its time-averaged force contribution $m v_{x}^{2} / L$ . Dividing again by the wall area $L^{2}$ gives a pressure contribution $m v_{x}^{2} / V$ . A slow molecule pushes less strongly and collides less often; both effects appear in the squared velocity. The derivation uses a time average for one molecule, then a sum over the molecular population. A pressure gauge effectively performs both averages over an enormous number of impacts.

For all molecules, the selected-wall result is $p V = \sum_{i} m v_{x i}^{2}$ . Random equilibrium motion has no preferred direction, so the sums of $v_{x}^{2}$ , $v_{y}^{2}$ , and $v_{z}^{2}$ are equal. Their sum is the total $v^{2}$ , yielding $p V = (1/3) N m \overline{v^{2}}$ . Momentum impulse and collision rate therefore make pressure depend on squared speed rather than average velocity. The vector average velocity of a stationary gas is zero, yet every wall receives positive momentum transfer. Bulk flow adds a nonzero average velocity and requires separating the ordered flow motion from the random thermal motion.

One molecule strikes the same wall only after a round trip across the container. Its average force combines the momentum impulse at impact with the interval between impacts.

Most probable, mean, and rms speeds

Three speed measures occur naturally in kinetic theory. The most probable speed $v_{mp} = 2 k T / m$ marks the peak of the speed distribution. The mean speed is the distribution-weighted average, $v_{av} = 8 k T / (π m)$ . The rms speed, $v_{rms} = 3 k T / m$ , is the square root of the average squared speed and is the speed measure that enters the pressure derivation. Their numerical order is $v_{mp} < v_{av} < v_{rms}$ . The high-speed tail shifts the mean above the peak and affects the rms speed most strongly because fast speeds are squared.

For nitrogen gas at $300 K$ , using molar mass $0.028 kg mo l^{- 1}$ , the three values are approximately $422 m s^{- 1}$ , $476 m s^{- 1}$ , and $517 m s^{- 1}$ , respectively. These are not three competing answers to one question. Each summarizes a different feature of the same distribution. A molecular-beam flux, a diffusion estimate, and a pressure calculation can call for different averages. Stating only that molecules move at the gas speed loses the information contained in the distribution.

At fixed temperature, all three scales vary as inverse square root of molecular mass. A helium atom consequently has a much larger characteristic speed than an oxygen molecule at the same temperature. At fixed mass, all three vary as square root of absolute temperature, not Celsius temperature. Doubling kelvin temperature increases each characteristic speed by a factor of $2$ , while quadrupling it doubles the speed scale. These proportionalities follow directly from the balance between molecular mass and average kinetic energy.

The three standard speed measures lie at different locations on one equilibrium distribution, with

v_{mp} < v_{av} < v_{rms}

. The rms speed is farthest right because high-speed molecules receive extra weight in the squared-speed average.

Translational degrees of freedom

A freely moving molecule has three independent translational velocity components: one along each Cartesian axis. These are the three translational degrees of freedom in the kinetic-theory model. Its translational kinetic energy is the sum $K_{tr} = (1/2) m v_{x}^{2} + (1/2) m v_{y}^{2} + (1/2) m v_{z}^{2}$ . Isotropy gives equal averages for the three squared components. Comparison with the ideal-gas law therefore gives average energy $(1/2) k T$ to each translational quadratic component and gives the total $(3/2) k T$ per molecule.

This restricted result is often called translational equipartition. It does not assert that every molecule has equal component energies at every instant. One molecule can temporarily carry most of its kinetic energy in the $x$ direction, while another moves mainly in the $z$ direction. The equality concerns the ensemble average after collisions have randomized directions. It also does not add rotational or vibrational energy storage to the present ideal translational model; those degrees require a broader molecular-energy treatment.

For $n$ moles, the total translational kinetic energy is $(3/2) n R T$ . The result is proportional to amount of gas and to absolute temperature. This molecular result explains why an ideal gas at fixed volume has pressure proportional to temperature: increasing temperature raises the average momentum transfer along each of the three directions. The factor three in the energy and the factor one third in the pressure derivation arise from the same three-dimensional motion.

The distinction among speed measures can be expressed directly as different weightings of the same distribution. The most probable speed is selected by the maximum of $f (v)$ . The mean speed uses the area-weighted quantity $\int_{0}^{\infty} v f (v) d v$ . The rms speed uses $\int_{0}^{\infty} v^{2} f (v) d v$ , which weights a molecule travelling twice as fast four times as strongly before the square root is taken. This weighting is why the rms value is linked to pressure: wall momentum transfer depends on squared normal speed after collision rate and impulse are combined.

The distribution is a speed distribution, not a distribution of one Cartesian velocity component. Molecules with the same speed can travel in many directions. Isotropy makes positive and negative components equally common, so the average of each velocity component is zero while the average of its square is positive. A pressure calculation uses these positive squared components. A net flow calculation uses the signed component average. Keeping these averages separate prevents the incorrect conclusion that a gas with zero bulk velocity has zero kinetic energy or zero pressure.

Probability intervals provide a practical way to use the curve. The fraction of molecules between two speeds is the area under the curve between the corresponding horizontal coordinates. A high-energy tail may occupy a small fraction of the area, yet it can control a process that has a threshold speed, such as escape from a weak gravitational field or passage over an activation barrier. Raising temperature enlarges that tail fraction even when the change in the most probable speed is modest. A single rms value is therefore sufficient for the ideal-gas pressure equation but insufficient for every molecular-rate question.

distribution quantity	mathematical weighting	information retained	appropriate use
most probable speed $v_{mp}$	location of the maximum of $f (v)$	the modal speed interval	Locate the peak of a measured speed histogram.
mean speed $v_{av}$	$\int_{0}^{\infty} v f (v) d v$	ordinary average of speed magnitude	Estimate a characteristic molecular transport speed when the relevant model calls for a first speed moment.
rms speed $v_{rms}$	$\int_{0}^{\infty} v^{2} f (v) d v$	squared-speed average, weighted toward the fast tail	Relate random molecular motion to pressure and translational kinetic energy.
component mean $\overline{v_{x}}$	signed average over positive and negative $v_{x}$	bulk motion along one axis	Determine flow velocity; it vanishes in an isotropic stationary gas.
tail fraction	$\int_{v_{*}}^{\infty} f (v) d v$	population above a specified threshold $v_{*}$	Predict a threshold-controlled molecular process rather than replacing the distribution by one representative speed.

Collision time and the pressure model.

The kinetic-theory derivation treats a wall collision as brief compared with the time between successive impacts. During the short contact interval, the wall force is large and the molecular momentum changes rapidly. During the much longer free flight, the idealized molecule experiences no force and carries constant momentum. The average force in the pressure calculation is the impulse from one brief event divided by the full interval from that event to the next collision with the same wall. Replacing the intermittent interaction by a steady average is legitimate only after specifying the time interval over which the pressure is observed.

The wall is assumed to be much more massive than one molecule and fixed in the laboratory frame. Under this approximation, the normal molecular velocity reverses with nearly unchanged magnitude, while tangential velocity is unchanged for a smooth wall. A moving piston or a rough wall changes the molecular energy as well as its momentum direction. The same momentum-transfer method still applies, but the collision rule must include the wall velocity or surface interaction. A constant-temperature ideal-gas description presumes that the container and gas have had sufficient interactions to reach thermal equilibrium.

Molecule--molecule collisions are not included explicitly in the wall-force sum, yet they are essential to the equilibrium assumptions. They redistribute speed and direction, preventing a persistent beam of molecules from crossing the container. The mean time between molecular collisions can be far shorter than the return time to a particular wall in a macroscopic vessel. That does not invalidate the pressure calculation: the sum uses the instantaneous distribution of velocity components at the wall, and equilibrium establishes the isotropic average required for the factor one third.

Pressure measurements also have a time-resolution limit. At very short time or area scales, the count of collisions fluctuates and the inferred force is noisy. Increasing the wall area or averaging time includes more impacts and narrows the relative fluctuation. A macroscopic gauge reports a stable pressure because it integrates over immense numbers of molecular events. In a rarefied gas or a microscopic chamber, collision statistics can no longer be hidden behind a smooth continuum pressure without estimating the event count.

Kinetic theory separates a brief impact from the much longer interval between impacts. Average force is the impulse divided by the repeat interval, while a real pressure reading averages over many intervals and molecules.

Mean free path and the dilute-gas limit

The mean free path is an average distance, not a fixed spacing between collisions. Some molecules travel only a short distance before meeting another molecule; others travel much farther through an unusually open region. For hard-sphere molecules of diameter $d$ and number density $n$ , the estimate $λ = 1/ (2 π d^{2} n)$ follows from a collision cross-section argument. A moving molecule sweeps out a cylindrical collision volume, and a collision is likely when another molecular centre lies within one diameter of its path. The relative motion of both collision partners produces the factor $2$ .

At ordinary atmospheric pressure, the mean free path is small on human scales but large compared with molecular size. Taking $n = 2.45 \times 1 0^{25} m^{- 3}$ and an effective diameter $d = 3.7 \times 1 0^{- 10} m$ gives roughly $6.7 \times 1 0^{- 8} m$ , or several dozen nanometres. This is hundreds of molecular diameters. The molecules are numerous, but most of the volume is empty space on the molecular scale. Decreasing pressure at fixed temperature decreases number density and increases the mean free path in inverse proportion.

The ideal-gas approximation is strongest when the molecular diameter is much less than the typical separation and when collision duration is short compared with free-flight time. The dimensionless density measure is $n d^{3}$ . When it is much less than one, excluded volume and sustained intermolecular forces are small corrections. Raising pressure or approaching condensation increases the importance of those corrections. The observed pressure can then differ from $nk T$ because repulsive volume exclusion and attractive forces alter the wall-collision picture.

Mean free path also determines whether a gas behaves as a continuum near a small object or channel. If the channel size is much larger than the mean free path, many collisions occur across it and pressure varies smoothly. If the sizes are comparable, molecules can travel from wall to wall with few intermolecular collisions. The pressure remains definable as a momentum flux, but standard continuum-flow assumptions need modification. This is a limit of the model's coarse-grained description, not a failure of momentum conservation.

A mean free path is the average distance a molecule covers between collisions in a dilute gas. Molecular diameters are small compared with the path length, which supports the point-particle and brief-collision approximations.

Inferring number density from pressure and temperature.

A constant-volume gas thermometer permits a direct molecular-scale inference. At fixed molecule number and volume, $p / T = (N / V) k = nk$ . A graph of absolute pressure against kelvin temperature is a straight line through the origin in the ideal limit. Its slope is $nk$ , so a measured pressure--temperature ratio determines number density without first weighing the gas. Celsius temperature cannot be used in this ratio because the line is not proportional to a shifted temperature scale.

At $p = 1.013 \times 1 0^{5} Pa$ and $T = 300 K$ , the ratio is $338 Pa K^{- 1}$ . Division by Boltzmann's constant gives $n = 2.45 \times 1 0^{25} m^{- 3}$ . A one-litre vessel at those conditions therefore contains about $2.45 \times 1 0^{22}$ molecules. The reciprocal number density gives volume per molecule; taking its cube root gives a characteristic centre-to-centre spacing of about $3.4 nm$ . This spacing is not a molecular diameter and not a mean free path. It is a density-derived length scale for a three-dimensional molecular arrangement.

The same measurement can infer average translational kinetic energy. The temperature obtained from the absolute pressure ratio gives $\overline{K}_{tr} = (3/2) k T$ , which is about $6.2 \times 1 0^{- 21} J$ per molecule at $300 K$ . Pressure alone does not determine this energy without number density or temperature; a dense cold gas and a sparse hot gas can have the same pressure. Measuring both pressure and temperature separates collision frequency from molecular kinetic-energy scale.

Experimental departures from a straight pressure--temperature line are diagnostic. A changing container volume, a gas leak, a nonuniform temperature, or nonideal molecular interactions can change the apparent slope. Before interpreting the slope as $nk$ , the apparatus must be allowed to equilibrate and the pressure must be absolute rather than gauge pressure. The molecular inference is only as reliable as those macroscopic controls.

At fixed volume and molecule number, absolute pressure is proportional to kelvin temperature. The slope is

nk

, so one pressure–temperature measurement fixes the number density and a molecular spacing scale.

A molecular-speed measurement model.

Molecular speeds can be inferred from a time-of-flight measurement when collisions along the flight path are rare. A short source pulse releases molecules into an evacuated tube of known length $L$ . A detector at the far end records the arrival time after the release trigger. For molecules travelling approximately along the tube axis, the speed estimate is $v = L / t$ . Repeating the pulse produces an arrival time distribution rather than one time, because the source contains molecules with many speeds. Early arrivals correspond to faster molecules; later arrivals sample the slower portion of the same distribution.

The measurement must distinguish flight time from instrument delay. If the source trigger and detector have a fixed electronic offset, it must be measured or subtracted before using $L / t$ . The tube length is the distance between the defined release and detection planes, not the outside length of the apparatus. A fractional uncertainty in length contributes directly to speed uncertainty, while a timing uncertainty contributes approximately $Δ t / t$ to the fractional speed uncertainty. Fast molecules have short flight times and can therefore be more sensitive to a fixed timing error.

An aperture selects directions as well as position. Without collimation, a molecule travelling at an angle follows a longer path than the axial separation, and the simple ratio underestimates its actual speed. A narrow source and detector acceptance angle reduce this geometric spread, at the cost of fewer detected molecules. A detector also has finite response time, so a very narrow arrival peak can be broadened by the instrument rather than by the gas. The measured histogram is consequently a convolution of the molecular distribution with source duration, path geometry, and detector response.

The gas must be sufficiently rarefied that a molecule normally reaches the detector without a collision. If the tube length is much greater than the mean free path, collisions randomize directions and times before detection, and the result no longer represents a simple source speed distribution. A pressure reduction increases the mean free path and restores ballistic transport. The required vacuum level depends on tube length and molecular diameter, connecting a laboratory measurement directly to the mean-free-path model.

A pulsed molecular beam converts flight time into a speed estimate

v = L / t

. Collimation fixes the path geometry, and a low pressure makes collisions inside the tube unlikely.

Conditions behind the ideal-gas approximation.

The ideal-gas equation is not a statement that molecules have no size or no interactions. It states that, under dilute conditions, those effects make small corrections to the pressure and energy relations. Molecular volume is negligible when the typical separation greatly exceeds the molecular diameter. Intermolecular forces are negligible for most of a free flight when molecules are far apart and act only briefly during a collision. Under those conditions, the container volume is nearly all available free volume and wall momentum transfer dominates the pressure description.

Two different departures appear as density rises. Finite molecular size reduces the volume available for translational motion, tending to make the measured pressure larger than a point-particle estimate at the same number density and temperature. Attractive forces can pull molecules inward before they reach a wall, reducing their wall momentum transfer and tending to reduce pressure. Which effect dominates depends on temperature and density. The simple relation $p = nk T$ cannot separate these mechanisms because it omits both from the model.

The ratio of mean free path to apparatus size is a separate issue from nonideal molecular forces. A very low-density gas can be close to ideal yet fail a continuum-flow approximation because molecules cross a narrow channel without many collisions. Conversely, a dense gas can behave as a smooth continuum while showing clear nonideal pressure corrections. One condition concerns the validity of treating molecules as independent point particles; the other concerns whether a spatially averaged pressure and velocity field describes a device.

Temperature matters because kinetic energy competes with intermolecular attraction. At high temperature, molecules pass through the attractive region rapidly and thermal motion dominates. At low temperature or near condensation, attraction becomes comparatively important and a gas can no longer be represented reliably by independent free flights. The warning signs are systematic curvature in a pressure--temperature plot at fixed volume, strong density dependence of the pressure ratio, or a measured compressibility inconsistent with the dilute model.

In the collision model, heating raises the average translational kinetic energy per molecule from $\frac{3}{2} k (300 K)$ to $\frac{3}{2} k (450 K)$ . The rms speed rises only by $450/300 = 1.225$ , not by $1.5$ ; pressure rises by $1.5$ because impulse per collision and collision frequency both grow with molecular speed, and together they scale with mean squared speed and hence with temperature.

The fixed-volume assumption is testable: a flexible vessel expands during heating and the pressure rise falls below the rigid-vessel prediction, and a leak lowers molecule number with the same qualitative effect. Several equilibrated pressure--temperature points distinguish a straight ideal trend from changes caused by volume, gas loss, or nonideal interactions.

Full microscopic derivation of the pressure equation

Take a cubical container with side length $L$ and select the wall perpendicular to the positive $x$ axis. Write the velocity of molecule $i$ as $(v_{x i}, v_{y i}, v_{z i})$ . Only $v_{x i}$ enters the momentum transferred to this particular wall. The tangential components can be nonzero, but a smooth elastic collision leaves them unchanged and they provide no normal impulse. Just before a collision with the selected wall, the normal molecular momentum is $m v_{x i}$ ; just after it, the normal momentum is $- m v_{x i}$ . Momentum conservation gives the wall an impulse $2 m v_{x i}$ directed outward from the gas.

A molecule moving toward that wall travels $2 L$ between successive collisions with the same wall. The recurrence interval is therefore $2 L / v_{x i}$ when the relevant velocity component is positive. Dividing impulse by recurrence interval gives the average normal force from that molecule, $m v_{x i}^{2} / L$ . Molecules that are currently travelling away from the selected wall give the same positive time-averaged contribution when their next round trip is included. Squaring the component removes any dependence on which side of the cube the molecule is moving toward at a particular instant.

Summing over the $N$ molecules gives the average force $F_{x} = (m / L) \sum_{i} v_{x i}^{2}$ . The wall area is $A = L^{2}$ , so division by area gives $p = (m / V) \sum_{i} v_{x i}^{2}$ . This expression is already a microscopic pressure equation, but it describes one wall direction. Equilibrium without a preferred direction gives the additional relation $\sum_{i} v_{x i}^{2} = \sum_{i} v_{y i}^{2} = \sum_{i} v_{z i}^{2}$ . Adding the three equal sums produces $\sum_{i} v_{i}^{2}$ , hence $p V = (1/3) m \sum_{i} v_{i}^{2} = (1/3) N m \overline{v^{2}}$ .

The final average is an average of squared speeds, not the square of average speed. A stationary sample has zero signed average velocity because every direction is equally represented. The average of $v^{2}$ remains positive and sets the pressure. This distinction is visible even for two molecules of equal speed moving in opposite directions: their vector momenta cancel, while their two wall impulses add. A pressure gauge measures momentum flux, not net momentum stored in the gas.

The dimensions check directly. Mass times squared speed has units $kg m^{2} s^{- 2}$ , which is energy. Division by volume gives $kg m^{- 1} s^{- 2}$ , the pascal. The relation also gives correct qualitative limits. At fixed volume and molecule number, doubling every molecular speed quadruples pressure. At fixed average squared speed, doubling molecule number doubles the collision rate and pressure. Both conclusions follow before inserting the ideal-gas temperature relation.

The derivation has several stated assumptions. The wall is rigid on molecular scales, collisions are elastic, the gas is spatially uniform, and molecular velocity directions are isotropic. A moving piston changes the collision impulse; a temperature gradient makes the velocity distribution position-dependent; a directed beam produces unequal component averages. The formula can be generalized in each case, but the simple scalar pressure relation then no longer follows from the symmetric cube argument alone.

Kinetic energy checks the rms value. The average translational energy at $300 K$ is $\frac{3}{2} k T = 6.21 \times 1 0^{- 21} J$ , and $\frac{1}{2} m v_{rms}^{2}$ with the calculated rms speed returns the same number. The check works because rms speed is defined from the average of $v^{2}$ ; substituting the mean speed into $\frac{1}{2} m v^{2}$ does not give the average kinetic energy, since the square must be averaged before the root is taken.

Heating the argon to $1200 K$ multiplies each characteristic speed by $1200/300 = 2$ , so the rms speed becomes $866 m s^{- 1}$ at unchanged mass. At fixed temperature, replacing argon with helium raises the speeds by the square root of the inverse mass ratio. The square-root scalings carry more than a memorized room-temperature value does.

The speed scale is much larger than familiar bulk flow speeds. A sealed room can contain air with zero average velocity while individual molecules move hundreds of metres per second. Collisions randomize direction so efficiently that the vector average vanishes. This separation between rapid molecular motion and slow or zero bulk motion is one reason a gas can exert a steady pressure on a stationary wall.

Assumptions and limits of Maxwell--Boltzmann statistics

The Maxwell--Boltzmann speed distribution describes a classical ideal gas in thermal equilibrium. Thermal equilibrium fixes one well-defined temperature and a time-independent distribution. Molecular collisions are frequent enough to randomize directions and redistribute energy, yet the gas is dilute enough that the molecules spend most of their time in free flight. The distribution is not a trajectory law for one molecule. A single molecule continually changes speed and direction through collisions; the curve describes the fraction of a very large population found in each speed interval at an instant or over a long average.

Random directions are as important as random speed magnitudes. The distribution of speeds is nonnegative and has a peak, whereas each Cartesian velocity component has positive and negative values and a symmetric distribution about zero. A gas with a nonzero bulk flow is best described by a thermal distribution around the flow velocity, not by treating the laboratory-frame speed distribution as isotropic without modification. In the rest frame of a uniform equilibrium sample, the three component variances are equal and the standard speed formulas apply directly.

The distribution function is normalized: adding the fractions in all possible speed intervals gives one. Temperature changes shift both the peak position and the width; one speed parameter cannot characterize the distribution. A hotter curve has more molecules in the high-speed tail even though many molecules remain slower than the rms speed. This tail is physically important whenever an outcome requires exceeding a speed threshold. A curve label at its peak cannot estimate a tail fraction; the relevant quantity is the area over the specified speed interval.

The ideal-gas form loses accuracy when molecular interactions alter free flights or when density makes excluded volume important. Near condensation, molecules are not independent between brief collisions and a simple Maxwell--Boltzmann speed description with $p = nk T$ is incomplete. Strong spatial gradients pose another limit. A gas next to a hot wall, a rapidly expanding jet, or a narrow rarefied channel can have different local distributions at different positions or a directionally biased distribution. One temperature and one isotropic curve then cannot represent the entire apparatus.

Statistical predictions also require a sufficient number of molecules in the sampled interval. In a macroscopic gas, the high-speed tail still contains an enormous count, so measured fractions are reproducible. In a tiny sample or a short-time detector count, the same tail can be represented by only a few events and fluctuate strongly. The distribution remains the expected ensemble result, but the observed histogram needs uncertainty estimates based on the number of detected molecules. Smooth curves summarize large-number behavior; they do not eliminate counting variation in a finite experiment.

Within its domain, Maxwell--Boltzmann statistics joins the pressure derivation to measurable molecular-speed data. The squared-speed average determines pressure and rms speed, the ordinary speed average determines a different transport scale, and integrated tail areas determine threshold probabilities. Each use selects a distinct average from the same distribution. The model should therefore be stated together with its equilibrium, dilute-gas, and frame assumptions rather than applied as a universal curve for every gas flow.

Three checks catch the usual errors in the derivation. Pressure must rise when molecular mass, collision frequency, or normal speed rises; reversing every molecular velocity must leave pressure unchanged, which forces squared components rather than a signed average; and the three Cartesian directions contribute equally only once isotropy is granted. The corresponding mistakes are a one-way crossing time (missing the round trip), a dropped momentum-reversal factor of two, and replacing $\overline{v^{2}}$ with $\overline{v}^{2}$ .