🔢 mathematics

Envelopes and evolutes

This paper defines the evolute of a projective variety as the envelope of its normal lines, characterizing it as a higher-order Thom-Boardman singularity to verify, generalize, and discover new enumerative formulas for the degrees and singularities of these geometric objects.

Original authors: Ragni Piene

Published 2026-03-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ragni Piene

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are walking along a winding path in a park. As you walk, you are constantly changing direction. If you were to draw a line sticking straight out from your back at every single step (a "normal line"), and then watched what happens when you move those lines just a tiny bit forward, they would start to crowd together and form a new, twisted shape.

This new shape is called an envelope. It's like the shadow cast by all those moving lines.

Now, imagine the most special points on that shadow—the sharp corners or the places where the lines cross over themselves. These special points form a new curve called the evolute.

This paper by Ragni Piene is a modern, high-level mathematical guide to understanding these shapes, but instead of just walking on a flat path in a park, the author is exploring them in the vast, abstract world of projective geometry (a kind of math that deals with shapes, lines, and spaces that can be curved, twisted, and exist in many dimensions at once).

Here is a breakdown of the paper's main ideas using simple analogies:

1. The Old Story vs. The New Map

The History:
Long ago, a man named Huygens (in the 1600s) figured out how to find the "center of curvature" for a flat curve. Think of it like finding the center of the circle that best fits a curve at any given moment. Later, other mathematicians like Monge and Salmon tried to do this for 3D objects and complex shapes. They came up with formulas to count how many "corners" or "twists" these shapes had, but their methods were often specific to certain types of shapes.

The New Approach:
Piene says, "Let's make a universal rulebook." Instead of just looking at flat curves or 3D surfaces, this paper creates a single, powerful mathematical machine that works for any shape in any number of dimensions. It uses a modern toolkit called Thom-Boardman singularities.

The Analogy: Imagine Salmon had a specific wrench for tightening bolts on a 19th-century car. Piene has invented a "universal robotic arm" that can tighten bolts on a 19th-century car, a modern spaceship, or a futuristic drone, all using the same logic.

2. The "Normal" Lines and the "Envelope"

To find these special shapes, the paper first needs to define what "straight out" (perpendicular) means in this abstract world.

The Setup: The author sets up a "Euclidean structure," which is just a fancy way of saying, "Let's pretend there is a standard way to measure right angles, even in this weird abstract space."
The Normal Bundle: Imagine a fence made of many sticks. At every point on a curve, you stick a pole straight out (perpendicular). The collection of all these poles is the "family of normal spaces."
The Envelope: If you wiggle those poles slightly, they will eventually touch and form a surface. This surface is the Envelope. It's the "skin" that wraps around all the possible positions of those normal lines.

3. The Evolute: The "Spine" of the Shape

The Evolute is the most interesting part. It's not the whole surface; it's the "spine" or the "backbone" of that envelope.

The Analogy: Think of a crumpled piece of paper (the envelope). The evolute is the sharp, jagged line where the paper folds over itself the most.
The Math Magic: The paper defines the evolute as a specific type of "singularity" (a point where the math gets messy or sharp). By using a known formula (a "Thom polynomial"), the author can calculate exactly how many of these sharp points exist without having to draw the shape first.

4. Counting the Corners (The Formulas)

The biggest achievement of the paper is that it gives us a way to count things.

For Curves: If you have a twisted wire in space, how many "vertices" (points of maximum curvature) does it have? The paper gives a formula to count them.
For Surfaces: If you have a complex 3D object (like a twisted balloon), how many "cusps" (sharp points) does its evolute have? The paper provides a formula for this too.

Why is this cool?
The author checks their new, high-tech formulas against the old formulas from Salmon (from the 1800s).

The Result: The new formulas match the old ones perfectly! But the new ones also work for shapes that Salmon never imagined (like 4D or 5D shapes). It's like verifying that your new GPS works on the old roads, but also shows you new shortcuts to places the old maps didn't even know existed.

5. The "Osculating" Developables

The paper also looks at "osculating" things.

The Analogy: "Osculating" means "kissing." An osculating circle is a circle that kisses a curve so perfectly that it matches the curve's bend at that exact moment.
The paper looks at what happens when you take these "kissing" shapes (planes, spheres, etc.) and wrap them up. It turns out these shapes are related to the envelopes and evolutes in a very neat, predictable way.

Summary: What did we learn?

This paper is a bridge between old-school geometry and modern algebraic topology.

It defines how to find the "spine" (evolute) of any shape in any dimension.
It uses a powerful counting tool (Thom polynomials) to predict how many sharp corners or special points these shapes will have.
It confirms that the old rules from the 19th century were correct, but now we have a much bigger, more flexible rulebook that works for the entire universe of mathematical shapes.

In short: The author took a complex, 100-year-old puzzle about the "centers" of curves and surfaces, solved it with a modern super-tool, and proved that the old solutions were right all along.

1. Problem Statement

The paper addresses the generalization of classical differential geometric concepts—specifically evolutes and envelopes—from real Euclidean spaces to complex projective varieties of arbitrary dimension.

Historical Context: The concept of the evolute (the locus of centers of curvature) for plane curves dates back to Huygens. It was extended to space curves by Monge (polar surfaces) and later studied by Salmon, who provided enumerative formulas for the degrees and singularities of evolutes for projective curves and surfaces.
The Gap: While classical results exist for specific dimensions (e.g., curves in $\mathbb{R}^2$ or $\mathbb{R}^3$ , surfaces in $\mathbb{R}^3$ ), there was no unified framework for defining evolutes for higher-dimensional varieties in projective space, nor a systematic method to compute their degrees and singularity classes using modern algebraic geometry tools.
Objective: To define evolutes for projective varieties of any dimension $r$ in $n$ -space, interpret them via singularity theory (Thom–Boardman singularities), and derive general enumerative formulas that recover and extend classical results (specifically those of Salmon).

2. Methodology

The author employs a combination of algebraic geometry, intersection theory, and singularity theory.

Definition of Envelopes:
- The paper defines the envelope of a family of linear spaces. Given a morphism $\psi: \mathbb{P}(F) \to \mathbb{P}(V)$ from the projectivization of a vector bundle $F$ over a variety $X$ to the ambient projective space, the envelope is defined as the branch locus of $\psi$ .
- From a singularity perspective, this envelope corresponds to the image of the Thom–Boardman singularity $\Sigma_1$ (where the rank of the differential drops).
Definition of Perpendicularity and Normal Spaces:
- To define "normal spaces" in projective space, the author introduces a Euclidean structure via a hyperplane at infinity $H_\infty$ and a non-singular quadric $Q_\infty \subset H_\infty$ .
- This allows for the definition of polar spaces and, consequently, normal spaces to a variety $X$ . The normal bundle $E$ is constructed as the dual of the conormal bundle (twisted by $\mathcal{O}(1)$ ) plus the trivial bundle.
Definition of Evolutes:
- The evolute of a variety $X$ of dimension $r$ in $n$ -space is defined as the image of the $(n-r)$ -th iterated cuspidal locus of the map from the total space of the normal spaces to the projective space.
- Mathematically, if the envelope is the $\Sigma_1$ locus, the evolute corresponds to the $\Sigma_{1, \dots, 1}$ locus (with $n-r$ ones). This interprets the evolute as a specific Thom–Boardman singularity.
Computational Tools:
- The paper utilizes Thom polynomials to compute the cohomology classes (and thus degrees) of these singular loci.
- The author derives explicit formulas for the Chern classes of the relevant bundles (tangent and normal bundles) and applies the projection formula to calculate degrees.

3. Key Contributions

A. Unified Definitions

Generalized Evolutes: The paper provides a rigorous definition of evolutes for any projective variety, not just curves or surfaces.
Iterated Cuspidal Loci: It establishes that evolutes are not arbitrary constructs but correspond to specific higher-order singularities ( $\Sigma_{1, \dots, 1}$ ) of the normal map.
Connection to Classical Objects: It clarifies that for curves in the plane, the evolute is the standard evolute; for space curves, it corresponds to the "cuspidal edge" of the polar surface; and for surfaces, it relates to the "surface of centers."

B. Enumerative Formulas

The author derives general formulas for the degrees of:

The Envelope ( $E_F$ ) of the normal spaces.
The Cuspidal Edge ( $C_F$ ) of the envelope (the first iterated singularity).
The Cuspidal Locus ( $\kappa_F$ ) of the cuspidal edge (the second iterated singularity).

These formulas are expressed in terms of:

Chern classes of the tangent bundle of $X$ ( $c_i(\Omega^1_X)$ ).
Chern classes of the normal bundle $F$ .
The degree of the variety and its genus (for curves).

C. Verification and Generalization of Salmon's Formulas

The paper rigorously verifies Salmon's classical formulas for curves in 3-space and surfaces in 3-space.
It resolves ambiguities in Salmon's work regarding inflection points and hyperosculating planes by introducing weighted counts ( $k_i$ ) derived from the theory of stationary indices.
It provides new formulas for higher-dimensional cases and for the degrees of higher-order singularities (e.g., the number of cusps on the evolute of a surface).

4. Key Results

For Curves in $\mathbb{P}^n$

Degree of the Envelope ($EE$):
$\deg EE = 6(d + g - 1) - 2k_0$
where $d$ is the degree, $g$ is the genus, and $k_0$ is the weighted number of cusps.
Degree of the Evolute ($CE$) (for $n=3$ ):
$\deg CE = 3(3d + 4g - 4 - k_0)$
This matches Salmon's formula $5m + \alpha$ when adjusted for inflection points.
Degree of the Cuspidal Locus ( $\kappa E$ ):
$\deg \kappa E = 4(3d + 5g - 5 - k_0)$
This corresponds to the number of vertices of the curve.

For Surfaces in $\mathbb{P}^3$

Degree of the Evolute (Surface of Centers):
$\deg EE = 2d(d-1)(2d-1)$
This confirms Salmon's result for a surface of degree $d$ .
Degree of the Cuspidal Curve on the Evolute:
$\deg CE = 2d(d-1)(11d - 16)$
Degree of the Cusps of the Cuspidal Curve:
$\deg \kappa E = 4d(30d^2 - 97d + 78)$
Umbilics: The paper discusses the number of umbilic points (where focal points coincide) and notes a discrepancy between the classical count and a naive application of Thom polynomials for $D_4$ singularities, suggesting a need for a more refined approach involving the $\Sigma_{2,2}$ locus.

For Osculating Developables

The paper generalizes the relationship between osculating developables and envelopes, proving that the envelope of osculating hyperplanes is the union of the $(n-2)$ -th osculating developable and the hyperosculating hyperplanes.

5. Significance

Bridging Classical and Modern Geometry: The paper successfully translates 19th-century enumerative geometry (Salmon, Monge) into the language of modern intersection theory and singularity theory (Thom–Boardman, Chern classes).
Computational Power: By utilizing known Thom polynomials, the author provides a systematic algorithm to compute degrees of evolutes and their singularities for varieties of any dimension, a task that was previously intractable or limited to low dimensions.
Conceptual Clarity: It clarifies the geometric nature of evolutes as specific singularity loci, linking the "centers of curvature" intuition to the rigorous definition of iterated cuspidal loci.
Foundation for Future Work: The framework established allows for the study of higher-order singularities in projective geometry and offers a template for analyzing other geometric envelopes defined by normal spaces.

In summary, Ragni Piene's paper provides a comprehensive algebraic-geometric framework for evolutes, unifying classical results with modern singularity theory and providing explicit, verifiable formulas for the degrees and singularities of these objects in projective space.