Carathéodory's criterion Created 2026-10-14 Updated 2026-10-31
A set is measurable for an outer measure when every set is split additively by :
The measurable sets form a sigma-algebra, and the restriction of to it is a complete measure.
Constantin Carathéodory Created 2026-10-18 Updated 2026-11-05
Constantin Carathéodory was a mathematician whose work connected measure theory, complex analysis and the calculus of variations. He began his career as an engineer and studied mathematics while working on a dam in Egypt, before entering mathematical research. His later contributions included links between the calculus of variations and partial differential equations, as well as conformal maps and their boundary behavior.[ref] His Caratheodory measurability criterion is a basic construction principle for measures: it selects those sets that split the outer measure of every test set additively, producing a domain on which countable additivity holds.
Figure 1. . Caratheodory family; photographer and date unknown, as stated in the Commons File. File records family permission as public domain and an origin-country/life-plus-70-or-shorter public-domain tag. Source usage terms, via Wikimedia Commons. Source.
Countable subadditivity of a measure Created 2026-10-05 Updated 2026-10-06
For measurable sets , a measure satisfies
Replace by the disjoint sets , use countable additivity, and then monotonicity. The same property is an axiom for an outer measure on arbitrary sets.
Hausdorff measure Created 2026-10-06 Updated 2026-10-25
Hausdorff measure measures the geometric size of a set using covers by increasingly small pieces. The exponent chooses the measuring scale: is length-like, is area-like, and noninteger exponents can measure fractals. Unlike a count of occupied grid cells, the construction allows pieces of different sizes and considers every admissible cover.
For and a subset of a metric space, use the diameter-cost convention
The first quantity is scale-restricted Hausdorff content. Shrinking removes options from the infimum, so these contents increase to a limit, possibly infinite. The resulting outer measure restricts to a measure on the Borel sets. For , counting each nonempty covering piece with cost one gives counting measure.
A different common convention multiplies these quantities by , using the Gamma function. At integer , that normalization agrees with ordinary Euclidean -dimensional volume. Here , so both conventions make agree with arc length on smooth curves in Euclidean space. Positive normalization constants never change Hausdorff dimension. A definition must specify this constant when numerical values of Hausdorff measure matter.
For a line segment of length , splitting it into equal pieces gives the cover cost
It tends to zero for , stays at for , and grows for . The matching lower bound comes from the fact that any interval cover must have total length at least . The exponent one is therefore the transition between a measuring rule too coarse to see the segment and one that assigns it infinite size. For a Cantor set, the transition is instead at , as shown in Hausdorff dimension of the Cantor set.
This makes Hausdorff measure useful when ordinary Lebesgue measure sees too little: a curve in the plane has zero planar area but can have positive arc length, and the Cantor set has zero line length but positive geometric size at its own dimension. It also separates local covering estimates from global cardinality: a countable dense set has zero Hausdorff dimension, despite meeting every open region of its closure.
An original source for the construction is Felix Hausdorff's [ref]. The modern covering and Hausdorff dimension framework is developed in Bishop and Peres, Fractals in Probability and Analysis, §1.2.
Lebesgue inner measure 2026-10-31
The Lebesgue inner measure approximates a subset of the real line from within by compact sets, while Lebesgue outer measure approximates it from outside by countable interval covers. For a finite closed interval containing ,
Indeed, every compact gives . Conversely, an open cover of with total length arbitrarily close to its outer measure has compact complement in lying inside , giving the reverse inequality in the limit.
For , equality of inner and outer measures holds precisely when is Lebesgue measurable. For an unbounded set, measurability is equivalently checked on all bounded intersections . Equality alone is insufficient: adjoining a bounded nonmeasurable subset to a disjoint half-line makes both quantities infinite without making the union measurable.
The mass distribution principle converts a diffuse measure into a lower bound on Hausdorff dimension. Let and let a Borel probability measure give a Borel set full mass. If, for some , every ball centered in satisfies
then in the diameter-cost convention, and consequently .
The idea is to keep unit mass from being crowded into cheap covering pieces. For a piece of positive diameter meeting , choose . A ball of radius contains ; letting gives , where is the associated outer measure. Zero-diameter pieces have zero mass by the same estimate. For any sufficiently small cover,
Thus every cover costs at least . This is the lower-bound counterpart to the explicit covers in scale-restricted Hausdorff content. The Cantor distribution supplies the needed mass in Hausdorff dimension of the Cantor set. [ref].
For a regular outer measure, its Caratheodory measurability criterion defines a maximal algebra of sets on which that outer measure is finitely additive. Indeed, suppose an additive algebra contains all measurable sets and also . Given of finite outer measure, take a measurable hull of equal mass. Finite additivity splits by , and monotonicity gives . Outer subadditivity gives equality. For infinite mass that equality already follows from outer subadditivity. Thus was measurable. No sigma-finite measure assumption is necessary.
Metric outer measure 2026-10-31
An outer measure on a metric space is metric when positively separated sets satisfy the displayed equality. Every Borel set is then measurable under Caratheodory measurability criterion. Lebesgue outer measure is metric: refine every box of a covering into finitely many boxes of diameter smaller than the separation. Each refined box meets at most one of the two sets, splitting the cover without changing its total volume. Taking infima proves the additive lower bound; outer subadditivity gives equality.
A ring of sets is closed under finite unions and differences. A premeasure on it is a nonnegative set function with value zero on the empty set that is countably additive whenever a disjoint union remains in the ring. Caratheodory extension theorem extends a premeasure to a measure on the generated sigma-algebra; this extension is unique there when the premeasure is sigma-finite. It is obtained from the outer measure
whose measurable sets satisfy for every . The resulting complete measurable sigma-algebra contains the sigma-algebra generated by the ring.
For the required continuity, let in the bounded-interval ring and fix . From each component of , remove a sufficiently short initial piece and retain the compact interval . The union is a compact set , and is covered by a ring set of total length less than ; one can take unions of . We do not incorrectly assume the compact sets themselves belong to the half-open interval ring.
Since and is compact, the finite intersection property supplies with . For , membership in all earlier means some failure of membership in puts in . Thus
Finite additivity gives the finite subadditivity used here. Monotonicity now proves .
If a disjoint union lies in the ring, its remainders decrease to the empty set. The proved continuity and finite additivity imply countable additivity, so the length function is a premeasure. The ring generates the Borel sigma-algebra and covers the line by bounded intervals of finite measure. Its unique sigma-finite extension is Borel length measure; its completion is Lebesgue measure.
Start with a finite subcover of the compact set . Among its remaining open balls, retain one of greatest radius and discard every ball intersecting it. Repeat until no balls remain. The retained family is finite and pairwise disjoint. If a discarded ball met the retained ball , then and . Therefore, for any ,
Every original ball lies in the concentric triple of one retained ball. This proves the Wiener covering lemma for the finite subcover. By Lebesgue measure scaling and subadditivity,
Disjointness gives
The empty compact set is covered by the empty selection.
The spherical derivative of a measure at is the limit
when it exists, where is the unit-ball volume. We prove that this limit is zero outside a Lebesgue-null set for the singular measure.
First derive the needed maximal estimate from the selection lemma. For a finite positive Borel measure , put
This is the uncentered maximal function of a finite measure. For , its strict superlevel set is the union of all balls with , so it is open. Cover any compact subset of this set by such balls and apply the finite Wiener covering lemma. The selected disjoint balls give
By inner regularity of Lebesgue measure, this proves the uncentered maximal weak-type inequality
Since and are mutually singular measures, there is a Borel set with and . By regularity of finite Borel measures on Euclidean space, for each choose compact with . Define the remainder measure .
For , distance from to this compact set is positive. Thus sufficiently small centered balls avoid , and
Write for the upper limit of the ratios. If it exceeds , then either or . Hence
The right side has Lebesgue measure at most , since . This even bounds the outer measure of the left side, without a separate measurability argument for the upper density. Let , and then take the countable union over positive rational . We obtain Lebesgue almost everywhere. The ratios are nonnegative, so their lower limit is also zero and
This proves that the spherical derivative of a singular measure vanishes. It is important to approximate the null carrier by compact subsets: a null carrier can be dense, so points outside it need not have any neighbourhood avoiding it.
In one dimension, consider the cumulative distribution function . For either sign of , monotonicity gives a nonnegative difference quotient, and the relevant half-open interval lies inside . In detail, the numerator is for , while after reversing both signs it is for . Thus
At each point where the spherical derivative of a measure is zero, this bound tends to zero from both sides. The enlarged open interval avoids any endpoint ambiguity from atoms. Therefore the ordinary two-sided derivative exists and
This is why singular distribution functions have zero derivative almost everywhere without having to be constant: almost-everywhere differentiation recovers increments only under extra hypotheses such as absolute continuity of a function.
The Lebesgue outer measure of is the infimum of over countable coverings by open intervals. A set is a null set when this infimum is zero, equivalently when it has such a cover of total length less than every prescribed positive number.
The Cantor set is obtained from by deleting the open middle third of every remaining interval at each stage. After stages it lies inside closed intervals of length . Slightly enlarging these finitely many intervals to open intervals shows its outer measure is at most for every , hence zero.
For the modified construction, let be the total length of the intervals in . The removed fraction at stage is exactly , so
The product telescopes because , giving
Since , the same finite interval-cover argument gives for every . Taking the infimum proves
Only interval lengths and the definition of outer measure were used; the three numbered clauses in the source describe the construction rather than separate requests to prove additional statements.
An outer measure on is a function satisfying , monotonicity, and countable subadditivity:
Its measurable sets are those satisfying Caratheodory measurability criterion:
The test runs over all subsets , including nonmeasurable ones.
A metric outer measure on a metric space is an outer measure satisfying
Here . Every Borel set is measurable for a metric outer measure. Consequently its measurable sigma-algebra contains the Borel sigma-algebra; it may be strictly larger.
For , define the Lebesgue outer measure by
where the boxes are bounded and . The empty covering handles the empty set. Monotonicity follows by inclusion of admissible covers. Taking covers within of the outer measures of a sequence of sets proves countable subadditivity; when the sum is infinite the inequality is automatic. Lebesgue measure is the restriction of to its Caratheodory-measurable sets.
If , subdivide each box of any countable cover of into finitely many half-open boxes of diameter smaller than , preserving its total volume. No small box can meet both and . The boxes meeting cover , and the disjoint family meeting covers , so the original cover costs at least . Taking the infimum and then using subadditivity gives
Thus the outer measure underlying Lebesgue measure is a metric outer measure, and in particular every Borel set is Lebesgue measurable.
A regular outer measure has, for every , a measurable hull with .
Suppose is an algebra containing and is finitely additive on . Fix and an arbitrary of finite outer measure. Take a measurable hull of . Both and belong to , so
Outer subadditivity gives equality. If , outer subadditivity already forces the sum on the right to be infinite, again giving equality. Therefore satisfies the Caratheodory criterion. We have proved , so
This proves the maximal additivity algebra of a regular outer measure without any sigma-finiteness assumption.
The Lebesgue outer measure is
where is the product of its side lengths. The infimum ranges over countable covers; allowing other standard endpoint conventions gives the same outer measure by arbitrarily small enlargement. Lebesgue measure is its restriction to Lebesgue measurable sets.
Enlarged unit cubes give . For any countable open-box cover of this compact cube choose a finite subcover. A finite grid formed from its box endpoints and the cube endpoints partitions the cube into rectangular cells. The sum of the volumes of covering boxes is at least the sum of these cell volumes, which is one. Boundary faces have elementary zero volume and do not affect the finite volume comparison. Thus every covering sum is at least one, proving
A set is Lebesgue measurable when the Caratheodory measurability criterion holds:
Subadditivity always gives the inequality with left side at most right side. For , split each covering box across the plane . Its two pieces have total volume . The piece with a closed cut face can be enlarged by arbitrarily little into an open box; hence its outer measure is at most its corresponding volume, just as for the other piece. Subadditivity applied separately on the two sides consequently gives
For finite outer measure take a covering sum arbitrarily close to ; infinite outer measure follows already from subadditivity. Thus every coordinate half space is measurable.
Carathéodory's theorem says that the measurable sets form a sigma-algebra and the outer measure restricted to it is a complete measure. Coordinate half spaces generate all open boxes and then all open sets, using countably many rational-endpoint boxes. They therefore generate the Borel sigma-algebra. Hence every Borel set is Lebesgue measurable.
A regular outer measure has a measurable hull of equal outer measure for every set: for each there is measurable with . Lebesgue outer measure has even a Borel hull. If its value on is finite, choose open-box covers of total cost below , and let be their open unions. Then is Borel and
For infinite outer measure the whole space is a suitable hull. Thus Lebesgue outer measure is regular.
If is measurable, additivity immediately gives the requested equality. Conversely suppose the equality holds. Choose a measurable hull of and put . Then and . Since is measurable, its splitting of gives
The assumed equality and imply . These subtractions are legitimate because . Outer-measure-null sets are measurable by the Caratheodory criterion. Hence is measurable, proving the equivalence.
Regular outer measure 2026-10-31
An outer measure is regular in this sense when every set admits a measurable hull of equal outer measure. This is different from merely specifying inner or outer approximation of already measurable sets by compact or open sets. Lebesgue outer measure has a hull that is a Borel set: for finite outer measure choose open covers of total volumes within of its value and intersect their unions. Monotonicity gives equality of the hull's outer measure with that of the original set. For infinite outer measure the whole Euclidean space is a hull. This regularity is useful for turning numerical equality of inner and outer sizes into measurability.