Laguerre–Forsyth Invariant¶

A cubic differential extracted from the normalized third-order equation of a nondegenerate projective plane curve.

Version: v1 · 2026-10-31 · History
Domain-specific #: 14527
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Projective Differential Geometry → Mathematics
Aliases: Laguerre-Forsyth invariant

Core Idea¶

The Laguerre–Forsyth invariant takes a nondegenerate local curve in the projective plane and assigns it a cubic differential.^[1] Its value is obtained from the coefficients of a normalized third-order equation satisfied by the curve's homogeneous coordinates.^[1] If that equation is written as \(z''' + 3P_2 z' + P_3 z=0\), the coefficient of the differential is \(P=P_3-\tfrac32 P_2'\), and the named invariant is \(P\,dt^3\).^[1] The differential, rather than an isolated coefficient written in one parameter, is the object the primary source identifies as invariant.^[1]

The construction is reusable across curves: choose an admissible local curve, obtain its normalized coordinate equation, and apply the same combination of coefficients. A quadratic curve supplies a boundary case for which \(P\) vanishes.^[1] Nonquadratic logarithmic spirals and exponential graphs, separately listed in the source, provide two settings in which the calculation can be nonzero.^[1] The explicit values below are calculations from the source's formula and curve families, not values quoted from the paper.

How would you explain it like I'm…

No faithful explanation at this level. All three agree a child-level picture turns the invariant into a single curviness score attached to a curve, whereas it is a cubic differential, not an isolated number.

Perspective-Proof Curve Measure

Mathematicians study curves in a kind of geometry where you're allowed to look at shapes through any slanted window, like how a shadow can stretch a circle into an oval. The Laguerre–Forsyth invariant is a measurement that runs along a curve, bit by bit, and stays meaningful no matter how you trace the curve. You get it from a special equation that the curve's position numbers obey. For curves like circles and ovals, the measurement is zero everywhere, while some spirals give something that isn't zero.

Projective Cubic Differential

In projective geometry, the Laguerre–Forsyth invariant attaches a quantity to a nondegenerate curve in the projective plane. You write the curve's homogeneous coordinates as solutions of a normalized third-order differential equation, z''' + 3P₂z' + P₃z = 0. From the coefficients you compute P = P₃ − (3/2)P₂', and the invariant is P dt³, a 'cubic differential'. The important point is that it's this differential, not P alone written in one particular parameter, that stays invariant. The same recipe works for any admissible curve. For a conic (quadratic curve), P is zero; for curves like non-quadratic logarithmic spirals or exponential graphs, it can be nonzero.

The Laguerre–Forsyth invariant assigns to a nondegenerate local curve in the projective plane a cubic differential. The homogeneous coordinates of the curve satisfy a third-order linear ODE, which can be normalized to the form z''' + 3P₂z' + P₃z = 0. From the coefficients one forms P = P₃ − (3/2)P₂', and the invariant is the differential P dt³. It is the differential, not the coefficient P expressed in one parameter, that the primary source identifies as invariant. The construction is a general recipe: choose an admissible local curve, find its normalized coordinate equation, and apply the same combination of coefficients. For a quadratic curve P vanishes, giving a boundary case, while nonquadratic logarithmic spirals and exponential graphs are settings where the result can be nonzero.

Structural Signature¶

Role sequence: nondegenerate local projective plane curve → three homogeneous-coordinate functions → normalized third-order equation → coefficient combination \(P_3-\tfrac32 P_2'\) → cubic differential \(P\,dt^3\).^[1]

Curve carrier. A local nondegenerate plane curve supplies three coordinate functions whose Wronskian is nonzero.^[1] This condition makes them a local solution triple for a third-order linear equation. If it fails, the stated construction cannot simply be applied to that triple.
Local parameter. A parameter \(t\) locates the coordinate equation and the displayed coefficient \(P(t)\); the output retains the \(dt^3\) factor.^[1] Changing the local parameter calls for treating the whole cubic differential as the invariant.
Normalization. Removing the second-derivative term gives the source's form \(z'''+3P_2z'+P_3z=0\).^[1] The functions \(P_2\) and \(P_3\) are intermediate data, not two separately named Laguerre–Forsyth invariants.^[1]
Combination. Differentiating \(P_2\) and subtracting \(\tfrac32P_2'\) from \(P_3\) is necessary.^[1] Using \(P_3\) alone, without subtracting \(\tfrac32P_2'\), changes the rule.^[1]
Output. The result is the cubic differential \(P\,dt^3\) associated with the curve.^[1] A displayed coefficient \(P(t)\) is always tied to the local parameter used in its computation.^[1]

The nondegeneracy and normalization steps are constitutive. Changing the sample curve changes the input; omitting those steps changes the computation itself.

What It Is Not¶

Not the curve. A logarithmic spiral or an exponential graph is an input on which the invariant can be evaluated. Neither particular curve defines the general identity.
Not the associated differential equation as a second example. The equation encodes the same curve for this calculation. Counting a curve and its equation as two settings would repeat one instance.
Not the raw coefficient \(P_3\). The defining combination includes the derivative of \(P_2\).^[1]
Not an unconditional global number. The source specifies the cubic differential in a local parameter. A coefficient reported at one parameter value does not by itself count global zeros or identify a curve everywhere.
Not the frozen article's unverified alternate formula. The expression \(P=(f')^2R\) and the equation \(x'''+Rx=0\) are withheld here because the inspected appendix states a different general presentation. The present entry makes no claim that those displays are equivalent.

Scope of Application¶

The literal carrier is a nondegenerate local projective plane curve for which the normalized coordinate equation is available.^[1] The source discusses projectively homogeneous curves and gives families on which the construction can be demonstrated.^[1] Its conic case shows what vanishing looks like within that carrier.^[1]

Local scope matters. For the exponential graph below, the chosen coordinate triple has Wronskian \((t+2)e^t\) and therefore works near \(t=0\); the same displayed normalization is not asserted across \(t=-2\).^[1] A point where the invariant vanishes, if present, is also a different statement from a curve for which it vanishes identically. The entry does not infer global counts, sextactic-point theorems, or a historical naming claim from the consulted appendix.

Clarity¶

The word invariant can conceal what is being preserved. Here the named object is a cubic differential, \(P\,dt^3\), obtained through the equation and its coefficients.^[1] It is useful to write the entire chain: curve, local parameter, normalized equation, \(P_2\), \(P_3\), and finally \(P\). That order prevents the coefficient \(P_3\) or a convenient unnormalized equation from being mistaken for the result.

The conic boundary is a test of the construction, not a definition by example. For \([1,\cos t,\sin t]\), the coordinate functions satisfy \(y'''+y'=0\).^[1] Thus \(P_2=1/3\), \(P_3=0\), and \(P=0\), agreeing with the source's quadratic-curve statement.^[1] A nonzero calculation in either positive example below shows that its local cubic differential differs from this vanishing case.

Manages Complexity¶

A projective plane curve admits different coordinate descriptions and local parameters. The invariant construction compresses that descriptive freedom into a small coefficient calculation while retaining the necessary conditions: nondegenerate triple, removal of the second-derivative term, and a cubic differential as output.^[1] The intermediate equation is a computational bridge, so one need not compare the raw coordinate lists directly to ask whether the named quantity vanishes locally.

That compression has a limit. A convenient numerical coefficient belongs to the parameter in which it was obtained. Comparing the bare numbers from unrelated parameterizations, or extending a local calculation through a zero of the Wronskian, would discard the information that makes the result well defined. The source's concise formula therefore saves work only when its carrier and normalization travel with it.

Abstract Reasoning¶

To test a proposed instance, first verify that the three coordinate functions are locally independent. Next derive their common third-order linear equation and eliminate its second-derivative term. Read off \(P_2\) and \(P_3\), differentiate \(P_2\), and calculate \(P_3-\tfrac32P_2'\). Finally report \(P\,dt^3\) and the local domain on which the derivation was valid.^[1]

This procedure separates three questions. Defined? The chosen curve and coordinate triple satisfy the local nondegeneracy requirement. Computed? The normalized coefficients yield the stated \(P\). Interpreted? A local nonzero value shows nonvanishing there; an identically zero calculation is stronger; neither assertion alone imports a global theorem about all zeros. Removing the derivative term from the formula would fail the computation question even if the same curve were used.

Knowledge Transfer¶

The method transfers between the source's different families because the same role sequence can be filled anew: the curve changes, but the coordinate equation, normalization, and cubic-differential formula stay in place.^[1] For a new plane curve, the first useful transfer is a diagnostic calculation of the Wronskian and then the normalized equation. The positive settings below demonstrate that a rotating exponential parametrization and a nonoscillatory graph can both supply admissible inputs.

The method does not license applying the Laguerre–Forsyth name to every differential invariant or every curve invariant. Another geometry or a degenerate coordinate triple needs its own defining construction. Nor does the invariant formula alone establish the frozen article's sextactic-point and global-count statements; those require their original theorems and hypotheses.

Examples¶

Canonical: a logarithmic spiral¶

The primary classification lists the logarithmic-spiral family \(X=e^{\lambda t}\cos(\mu t)\), \(Y=e^{\lambda t}\sin(\mu t)\).^[1] Choose \(\lambda=\mu=1\) and use homogeneous coordinates \([1,e^t\cos t,e^t\sin t]\).^[1] Direct differentiation gives Wronskian \(2e^{2t}\), so this triple is nondegenerate.^[1] The coordinates satisfy \(y'''-2y''+2y'=0\).^[1] With \(z=e^{-2t/3}y\), the normalized equation gives \(P_2=2/9\) and \(P_3=20/27\), hence \(P=20/27\) and \(P\,dt^3\) is nonzero.^[1]

Mapped back: spiral → nondegenerate coordinate triple → normalized equation → \(P_3-\tfrac32P_2'=20/27\) → nonzero local cubic differential.^[1] The exact value is a derivation using the paper's equation and family, rather than a numerical assertion printed there.

Applied: an exponential graph¶

The same classification separately lists \(Y=Xe^X\).^[1] On a neighborhood of \(t=0\), use \([1,t,te^t]\).^[1] Direct differentiation gives Wronskian \((t+2)e^t\), which is nonzero near zero.^[1] Its coordinates satisfy \(y'''-\frac{t+3}{t+2}y''=0\).^[1] The normalized lift \(z=e^{-t/3}(t+2)^{-1/3}y\) gives \(P_2=-\frac{t^2+6t+12}{9(t+2)^2}\) and \(P_3=-\frac{2(t^3+9t^2+27t+18)}{27(t+2)^3}\).^[1] In particular, \(P(0)=-5/12\), so the cubic differential is nonzero near that point.^[1]

Mapped back: exponential graph → locally nondegenerate coordinate triple → normalized variable-coefficient equation → \(P_3-\tfrac32P_2'\) → nonzero local cubic differential.^[1] The graph and spiral are separately enumerated source families, not one curve paired with its associated equation.^[1] Their coordinate shapes and the displayed coefficient behavior are illustrative; those displays alone are not a proof of projective inequivalence.

Structural Tensions¶

T1: Coordinate convenience versus invariant meaning. A simple parameter makes the spiral calculation short, but the meaningful object is \(P\,dt^3\).^[1] Diagnostic: Is a coefficient being interpreted without its local parameter and differential weight?

T2: Local calculation versus global claim. The graph calculation proves nonvanishing near \(t=0\), while a theorem about all zeros on a closed curve needs more hypotheses and evidence. Diagnostic: Does the conclusion say only what the calculated neighborhood supports?

T3: Compact formula versus admissible input. The expression \(P_3-\tfrac32P_2'\) is easy to copy, but its coefficients come from a normalized equation for a nondegenerate curve.^[1] Diagnostic: Were nondegeneracy and normalization actually checked?

T4: Named invariant versus neighboring descriptions. The curve, its coordinate equation, an osculating construction, and a general differential invariant can be related without being the same identity. Diagnostic: Which object has the stated cubic-differential definition?

T5: Autonomous identity versus reduction to a generic invariant. The result exemplifies preservation under a projective change, but that broad description does not recover its curve carrier, coefficient correction, or cubic weight. The narrow formula does not erase the broader invariance principle either. Diagnostic: Does a proposed broader concept contain the entire named construction, or only its preservation theme?

Structural–Framed Character¶

Vocabulary portability: invariance and normalization are broad mathematical ideas, but \(P_3-\tfrac32P_2'\) from this curve equation is specific. Evaluative weight: vanishing or nonvanishing is a mathematical property, not a judgment about the curve. Institutional origin: the historical name identifies a construction; no institution is part of its input. Human-practice dependence: choosing coordinates is a human calculation, while the invariant is intended to describe the curve independently of that display. Import versus recognition: a general curve invariant is not this invariant unless it follows the named coefficient construction. Its character: predominantly structural within a specialized mathematical frame: its identification turns on a checkable invariant construction, while the projective-curve carrier and cubic differential keep it domain-specific.

Structural Core vs. Domain Accent¶

The portable skeleton is normalization followed by an invariant combination of coefficients. Its domain-defining content is a nondegenerate projective plane curve, a third-order coordinate equation, the exact derivative correction \(P_3-\tfrac32P_2'\), and cubic differential \(P\,dt^3\).^[1] Removing those details yields a general method of constructing invariants, not the Laguerre–Forsyth invariant.

Future-Prime question: Is there a sourced recurrence in another domain with the same necessary roles and an all-instance rule that survives comparison against existing Primes, beyond the analogy of combining coefficients to remove coordinate dependence? This article establishes only the projective-curve construction. It therefore proposes a domain-specific entry while leaving any cross-domain Prime claim unmade.

This entry presupposes Curve and is a kind of Differential invariant.

The live Differential Invariant entry is the strict genus parent: this cubic differential has a coefficient \(P\) that is a weight-three relative function of a finite curve jet, while \(P\,dt^3\) remains the geometric output.^[2] Doubrov and Dunajski give the same coefficient combination, an explicit graph formula through the fifth derivative, and the prolonged projective-group action on curve jets.^[2] The child retains its nondegenerate projective plane-curve carrier and exact correction formula, so it is narrower than that genus. The live Curve entry is a separate prerequisite: without a curve there is no coordinate equation from which to derive the invariant.^[1] The Invariance Prime is a broader ancestor through Differential Invariant, while Osculating Curve remains a neighboring construction with a different output. The isolated official graph check confirmed both typed relations without a cycle.

Relationships to Other Abstractions¶

Current abstraction Laguerre–Forsyth Invariant Domain-specific

Parents (2) — outgoing typed relations

Laguerre–Forsyth Invariant is a kind of Differential invariant Domain-specific

The cubic form has a weight-three relative finite-jet representative under the projective group action.
Laguerre–Forsyth Invariant presupposes Curve Domain-specific

A nondegenerate local projective plane curve is required to supply the invariant's coordinate equation.

Hierarchy paths (3) — routes to 2 parentless roots

Laguerre–Forsyth Invariant → is a kind of Differential invariant → is a kind of Invariance

Show alternative paths (2)

Neighborhood in Abstraction Space¶

Laguerre–Forsyth Invariant sits in a sparse region of the domain-specific corpus (90^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2835 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-31

Not to Be Confused With¶

The Laguerre–Forsyth invariant is neither a generic scalar curvature nor the source's entire classification of homogeneous curves. It is also distinct from the Laguerre formula in another mathematical context. A zero value on the conic example should not be confused with a failure to define the invariant; the nondegenerate conic still admits the construction. Conversely, a degenerate triple cannot be treated as an ordinary zero result. The frozen source's sextactic and homotopy assertions remain separate research questions rather than conclusions of this entry.

References¶

[1] Franki Dillen, Takeshi Sasaki, and Luc Vrancken, “The Classification of Projectively Homogeneous Surfaces II,” Osaka Journal of Mathematics 35, no. 1 (1998): 117–146, Appendix §1, 143–144, equations (a.1)–(a.3), curve families [P1]–[P3], and Proposition A1. https://doi.org/10.18910/11131. Open published PDF. The numerical examples here are independent symbolic calculations from the paper's formula and listed families, not quoted results. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x ↩y ↩z ↩27 ↩28 ↩29 ↩30 ↩31 ↩32 ↩33 ↩34 ↩35 ↩36 ↩37 ↩38 ↩39 ↩40 ↩41 ↩42 ↩43 ↩44

[2] Boris Doubrov and Maciej Dunajski, “Co-calibrated G2 Structure from Cuspidal Cubics,” Annals of Global Analysis and Geometry 42 (2012): 247–265, §5, equations (5.21)–(5.23) and footnote 3. https://doi.org/10.1007/s10455-012-9310-8. Published author copy. Consulted the local curve-jet formula, relative weight-three coefficient, invariant cubic differential and prolonged projective-group action, not the paper's global cuspidal-cubic classification. registry ↩a ↩b