Introduction

Throughout classical geometry, we measure objects using:

• Length
• Area
• Volume

For example:

A line segment has length.

A square has area.

A cube has volume.

These measurements correspond to dimensions:

Object

Dimension

Line

1

Surface

2

Solid

3

For centuries mathematicians assumed that dimension must be an integer.

Then objects like the:

Cantor Set

appeared.

The Cantor set is:

• larger than a finite collection of points
• smaller than an interval
• dimension neither 0 nor 1

This raised a profound question:

How do we measure objects whose dimension is not an integer?

The answer is Hausdorff Measure.

Hausdorff measure is one of the most important ideas in modern geometry.

It allows us to measure:

• Fractals
• Singular sets
• Irregular geometric objects
• Infinite-dimensional phenomena

and forms the foundation of geometric measure theory.

---

Why Lebesgue Measure Is Not Enough

Consider the Cantor set:

$$C$$

We proved earlier that:

$$\lambda(C)=0$$

Thus Lebesgue measure says:

The Cantor set has no length.

But the Cantor set is certainly not empty.

It is uncountable.

In fact:

$$|C|=|\mathbb R|$$

Lebesgue measure loses important geometric information.

We need a more refined notion of size.

---

The Basic Idea

Recall how length is measured.

Cover a set by intervals.

Add the interval lengths.

Take limits.

Hausdorff asked:

What if we replace interval length by diameter raised to an arbitrary power?

This simple idea changes everything.

---

Diameter of a Set

The diameter of a set:

$$A$$

is:

$$\operatorname{diam}(A)=\sup{|x-y|:x,y\in A}$$

Intuitively:

Diameter measures the largest distance between points in the set.

---

Examples

For:

$$[0,1]$$

the diameter is:

$$1$$

For:

$$[2,5]$$

the diameter is:

$$3$$

For a single point:

$${x}$$

the diameter is:

$$0$$

---

Coverings

Suppose:

$$E\subseteq\mathbb R^n$$

A collection:

$${U_i}$$

is called a cover if:

$$E\subseteq\bigcup_i U_i$$

We imagine covering:

$$E$$

using many tiny sets.

---

Introducing Dimension s

Fix:

$$s\ge0$$

Instead of summing lengths,

we sum:

$$\operatorname{diam}(U_i)^s$$

For a given cover:

$$\sum_i \operatorname{diam}(U_i)^s$$

Different values of:

$$s$$

produce different notions of size.

---

Hausdorff Outer Measure

Restrict attention to covers whose diameters satisfy:

$$\operatorname{diam}(U_i)<\delta$$

Define:

$$H_\delta^s(E)=\inf\left{\sum_i \operatorname{diam}(U_i)^s\right}$$

where the infimum is taken over all such covers.

Now let:

$$\delta\to0$$

The resulting limit is the:

s-dimensional Hausdorff measure

written:

$$H^s(E)$$

---

Why This Definition Is Brilliant

The parameter:

$$s$$

acts like a dimension.

Different choices of:

$$s$$

probe different geometric scales.

---

Example: An Interval

Consider:

$$[0,1]$$

When:

$$s=1$$

Hausdorff measure essentially reproduces length:

$$H^1([0,1])=1$$

Thus ordinary length appears naturally.

---

Example: A Square

For:

$$[0,1]^2$$

we obtain:

$$H^2([0,1]^2)=1$$

This corresponds to area.

---

Example: A Cube

For:

$$[0,1]^3$$

we obtain:

$$H^3([0,1]^3)=1$$

This corresponds to volume.

---

A Remarkable Fact

Hausdorff measure generalizes all ordinary geometric measurements.

Dimension

Hausdorff Measure

1

Length

2

Area

3

Volume

The classical world becomes a special case.

---

What Happens for the Cantor Set?

Recall the Cantor set:

$$C$$

At each stage:

• number of pieces doubles
• lengths shrink by:

$$\frac13$$

This suggests the critical dimension should satisfy:

$$2\left(\frac13\right)^s=1$$

Solving:

$$s=\frac{\log 2}{\log 3}$$

This value is approximately:

$$0.63093$$

---

A New Kind of Dimension

The number:

$$\frac{\log 2}{\log 3}$$

is neither:

$$0$$

nor:

$$1$$

It is fractional.

This is the first appearance of a fractal dimension.

---

Critical Behavior

A remarkable theorem states:

For the Cantor set:

$$H^s(C)=\infty$$

when:

$$s<\frac{\log2}{\log3}$$

and:

$$H^s(C)=0$$

when:

$$s>\frac{\log2}{\log3}$$

At the critical dimension:

$$s=\frac{\log2}{\log3}$$

the measure is finite and positive.

This phenomenon occurs throughout fractal geometry.

---

Hausdorff Dimension

The critical value where the transition occurs is called the Hausdorff dimension.

Formally:

$$\dim_H(E)=\inf{s:H^s(E)=0}$$

Equivalently:

$$\dim_H(E)=\sup{s:H^s(E)=\infty}$$

---

Examples

A point:

$$\dim_H({x})=0$$

A line segment:

$$\dim_H([0,1])=1$$

A square:

$$\dim_H([0,1]^2)=2$$

The Cantor set:

$$\dim_H(C)=\frac{\log2}{\log3}$$

---

Why This Is Revolutionary

Dimension is no longer restricted to integers.

Objects can possess dimensions such as:

$$0.63$$

$$1.26$$

$$2.71$$

Geometry suddenly becomes far richer.

---

Example: Coastlines

Real coastlines are not smooth curves.

Their measured length depends on the scale used.

Hausdorff dimension provides a way to quantify this complexity.

This insight helped launch modern fractal geometry.

---

Example: Brownian Motion

A Brownian path is highly irregular.

Its Hausdorff dimension is:

$$2$$

even though it is a curve.

This surprising fact plays an important role in probability theory.

---

Hausdorff Measure and Singular Measures

Many singular measures live on sets of fractional dimension.

Examples include:

• Cantor measure
• Self-similar measures
• Fractal probability distributions

Hausdorff measure provides the natural geometric language for studying them.

---

Why Analysts Care

Hausdorff measure appears in:

• Fractal geometry
• Harmonic analysis
• PDEs
• Dynamical systems
• Probability theory

Many modern research areas rely heavily on it.

---

Connection to Geometric Measure Theory

Classical geometry studies smooth objects.

Geometric measure theory studies:

• rough sets
• singular surfaces
• fractals

Hausdorff measure is the foundational tool of the subject.

---

Connection to Probability

Random fractals often possess non-integer dimensions.

Examples include:

• Brownian motion
• Branching processes
• Percolation clusters

Hausdorff dimension allows these random objects to be studied quantitatively.

---

Philosophical Meaning

Hausdorff measure teaches an important lesson:

Size and dimension are not the same thing.

A set may:

• have measure zero
• be uncountable
• possess a non-integer dimension

The Cantor set demonstrates all three simultaneously.

---

Connection to Alain Connes

One of the motivations behind noncommutative geometry is that classical geometry is too restrictive.

Fractals already demonstrate this limitation.

Objects such as the Cantor set cannot be adequately described using ordinary geometric intuition.

Connes’ work extends geometry even further by creating tools for spaces that may not possess points in the usual sense.

Hausdorff measure is therefore one of the first major steps away from classical Euclidean geometry and toward the broader geometric worldview that eventually culminates in noncommutative geometry.

---

Key Concepts Learned

By the end of this lesson you should understand:

• Hausdorff measure generalizes length, area, and volume.
• The diameter of a set is:

$$\operatorname{diam}(A)=\sup{|x-y|:x,y\in A}$$

• Hausdorff measure depends on a parameter:

$$s$$

interpreted as dimension.

• Hausdorff dimension is the critical value where Hausdorff measure changes from infinity to zero.
• The Cantor set has dimension:

$$\frac{\log2}{\log3}$$

• Hausdorff measure is fundamental in fractal geometry.
• Many singular measures live on sets with fractional dimension.

---

Looking Ahead

Measure Theory Lesson 34: Fractal Measures and Self-Similarity

In the next lesson, we will move beyond measuring fractals and begin studying measures that naturally live on fractals. We will investigate self-similar measures, Cantor-type measures, scaling laws, and the emergence of dimension from repeated geometric patterns. These ideas form an important bridge between measure theory, probability, dynamical systems, and the geometric ideas that later influence noncommutative geometry.