History of Lorentz transformations

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

The history of Lorentz transformations comprises the development of linear transformations forming the Lorentz group or Poincaré group preserving the Lorentz interval and the Minkowski inner product .

In mathematics, transformations equivalent to what was later known as Lorentz transformations in various dimensions were discussed in the 19th century in relation to the theory of quadratic forms, hyperbolic geometry, Möbius geometry, and sphere geometry, which is connected to the fact that the group of motions in hyperbolic space, the Möbius group or projective special linear group, and the Laguerre group are isomorphic to the Lorentz group.

In physics, Lorentz transformations became known at the beginning of the 20th century, when it was discovered that they exhibit the symmetry of Maxwell's equations. Subsequently, they became fundamental to all of physics, because they formed the basis of special relativity in which they exhibit the symmetry of Minkowski spacetime, making the speed of light invariant between different inertial frames. They relate the spacetime coordinates of two arbitrary inertial frames of reference with constant relative speed v. In one frame, the position of an event is given by x,y,z and time t, while in the other frame the same event has coordinates x′,y′,z′ and t′.

Most general Lorentz transformations[edit]

The general quadratic form q(x) with coefficients of a symmetric matrix A, the associated bilinear form b(x,y), and the linear transformations of q(x) and b(x,y) into q(x′) and b(x′,y′) using the transformation matrix g, can be written as[1]






in which case n=1 is the binary quadratic form, n=2 is the ternary quadratic form, n=3 is the quaternary quadratic form.

Learning materials from Wikiversity: The binary quadratic form was introduced by Lagrange (1773) and Gauss (1798/1801), and the ternary quadratic form by Gauss (1798/1801).

The general Lorentz transformation follows from (Q1) by setting A=A′=diag(-1,1,...,1) and det g=±1. It forms an indefinite orthogonal group called the Lorentz group O(1,n), while the case det g=+1 forms the restricted Lorentz group SO(1,n). The quadratic form q(x) becomes the Lorentz interval in terms of an indefinite quadratic form of Minkowski space (being a special case of pseudo-Euclidean space), and the associated bilinear form b(x) becomes the Minkowski inner product:[2][3]






Learning materials from Wikiversity: Such general Lorentz transformations (1a) for various dimensions were used by Gauss (1818), Jacobi (1827, 1833), Lebesgue (1837), Bour (1856), Somov (1863), Hill (1882) in order to simplify computations of elliptic functions and integrals.[4][5] They were also used by Poincaré (1881), Cox (1881/82), Picard (1882, 1884), Killing (1885, 1893), Gérard (1892), Hausdorff (1899), Woods (1901, 1903), Liebmann (1904/05) to describe hyperbolic motions (i.e. rigid motions in the hyperbolic plane or hyperbolic space), which were expressed in terms of Weierstrass coordinates of the hyperboloid model satisfying the relation or in terms of the Cayley–Klein metric of projective geometry using the "absolute" form .[6][7] In addition, infinitesimal transformations related to the Lie algebra of the group of hyperbolic motions were given in terms of Weierstrass coordinates by Killing (1888-1897).

If in (1a) are interpreted as homogeneous coordinates, then the corresponding inhomogenous coordinates follow by

so that the Lorentz transformation becomes a homography leaving invariant the equation of the unit sphere, which John Lighton Synge called "the most general formula for the composition of velocities" in terms of special relativity (the transformation matrix g stays the same as in (1a)):[8]






Learning materials from Wikiversity: Such Lorentz transformations for various dimensions were used by Gauss (1818), Jacobi (1827–1833), Lebesgue (1837), Bour (1856), Somov (1863), Hill (1882), Callandreau (1885) in order to simplify computations of elliptic functions and integrals, by Picard (1882-1884) in relation to Hermitian quadratic forms, or by Woods (1901, 1903) in terms of the Beltrami–Klein model of hyperbolic geometry. In addition, infinitesimal transformations in terms of the Lie algebra of the group of hyperbolic motions leaving invariant the unit sphere were given by Lie (1885-1893) and Werner (1889) and Killing (1888-1897).

Lorentz transformation via imaginary orthogonal transformation[edit]

By using the imaginary quantities in x as well as (s=1,2...n) in g, the Lorentz transformation (1a) assumes the form of an orthogonal transformation of Euclidean space forming the orthogonal group O(n) if det g=±1 or the special orthogonal group SO(n) if det g=+1, the Lorentz interval becomes the Euclidean norm, and the Minkowski inner product becomes the dot product:[9]






Learning materials from Wikiversity: The cases n=1,2,3,4 of orthogonal transformations in terms of real coordinates were discussed by Euler (1771) and in n dimensions by Cauchy (1829). The case in which one of these coordinates is imaginary and the other ones remain real was alluded to by Lie (1871) in terms of spheres with imaginary radius, while the interpretation of the imaginary coordinate as being related to the dimension of time as well as the explicit formulation of Lorentz transformations with n=3 was given by Minkowski (1907) and Sommerfeld (1909).

A well known example of this orthogonal transformation is spatial rotation in terms of trigonometric functions, which become Lorentz transformations by using an imaginary angle , so that trigonometric functions become equivalent to hyperbolic functions:






or in exponential form using Euler's formula :






Learning materials from Wikiversity: Defining as real, spatial rotation in the form (2b-1) was introduced by Euler (1771) and in the form (2c-1) by Wessel (1799). The interpretation of (2b) as Lorentz boost (i.e. Lorentz transformation without spatial rotation) in which correspond to the imaginary quantities was given by Minkowski (1907) and Sommerfeld (1909). As shown in the next section using hyperbolic functions, (2b) becomes (3b) while (2c) becomes (3d).

Lorentz transformation via hyperbolic functions[edit]

The case of a Lorentz transformation without spatial rotation is called a Lorentz boost. The simplest case can be given, for instance, by setting n=1 in (1a):

or in matrix notation






which resembles precisely the relations of hyperbolic functions in terms of hyperbolic angle . Thus by adding an unchanged -axis, a Lorentz boost or hyperbolic rotation for n=2 (being the same as a rotation around an imaginary angle in (2b) or a translation in the hyperbolic plane in terms of the hyperboloid model) is given by

or in matrix notation






in which the rapidity can be composed of arbitrary many rapidities as per the angle sum laws of hyperbolic sines and cosines, so that one hyperbolic rotation can represent the sum of many other hyperbolic rotations, analogous to the relation between angle sum laws of circular trigonometry and spatial rotations. Alternatively, the hyperbolic angle sum laws themselves can be interpreted as Lorentz boosts, as demonstrated by using the parameterization of the unit hyperbola:

or in matrix notation






Finally, Lorentz boost (3b) assumes a simple form by using squeeze mappings in analogy to Euler's formula in (2c):[10]






Learning materials from Wikiversity: Hyperbolic relations (a,b) on the right of (3b) were given by Riccati (1757), relations (a,b,c,d,e,f) by Lambert (1768–1770). Lorentz transformations (3b) were given by Laisant (1874), Cox (1882), Lindemann (1890/91), Gérard (1892), Killing (1893, 1897/98), Whitehead (1897/98), Woods (1903/05) and Liebmann (1904/05) in terms of Weierstrass coordinates of the hyperboloid model. Hyperbolic angle sum laws equivalent to Lorentz boost (3c) were given by Riccati (1757) and Lambert (1768–1770), while the matrix representation was given by Glaisher (1878) and Günther (1880/81). Lorentz transformations (3d-1) were given by Lindemann (1890/91) and Herglotz (1909), while formulas equivalent to (3d-2) by Klein (1871).

In line with equation (1b) one can use coordinates inside the unit circle , thus the corresponding Lorentz transformations (3b) obtain the form:






Learning materials from Wikiversity: These Lorentz transformations were given by Escherich (1874) and Killing (1898) (on the left), as well as Beltrami (1868) and Schur (1885/86, 1900/02) (on the right) in terms of Beltrami coordinates[11] of hyperbolic geometry.

By using the scalar product of , the resulting Lorentz transformation can be seen as equivalent to the hyperbolic law of cosines:[12][R 1][13]






Learning materials from Wikiversity: The hyperbolic law of cosines (a) was given by Taurinus (1826) and Lobachevsky (1829/30) and others, while variant (b) was given by Schur (1900/02).

Lorentz transformation via velocity[edit]

In the theory of relativity, Lorentz transformations exhibit the symmetry of Minkowski spacetime by using a constant c as the speed of light, and a parameter v as the relative velocity between two inertial reference frames. In particular, the hyperbolic angle in (3b) can be interpreted as the velocity related rapidity , so that is the Lorentz factor, the proper velocity, the velocity of another object, the velocity-addition formula, thus (3b) becomes:






Or in four dimensions and by setting and adding an unchanged z the familiar form follows, using as Doppler factor:






In physics, analogous transformations have been introduced by Voigt (1887) and by Lorentz (1892, 1895) who analyzed Maxwell's equations, they were completed by Larmor (1897, 1900) and Lorentz (1899, 1904), and brought into their modern form by Poincaré (1905) who gave the transformation the name of Lorentz.[14] Eventually, Einstein (1905) showed in his development of special relativity that the transformations follow from the principle of relativity and constant light speed alone by modifying the traditional concepts of space and time, without requiring a mechanical aether in contradistinction to Lorentz and Poincaré.[15] Minkowski (1907–1908) used them to argue that space and time are inseparably connected as spacetime. Minkowski (1907–1908) and Varićak (1910) showed the relation to imaginary and hyperbolic functions. Important contributions to the mathematical understanding of the Lorentz transformation were also made by other authors such as Herglotz (1909–10), Ignatowski (1910), Noether (1910) and Klein (1910), Borel (1913–14).

Learning materials from Wikiversity: In pure mathematics, similar transformations have been used by Lipschitz (1885/86).

Also Lorentz boosts for arbitrary directions in line with (1a) can be given as:[16]

or in vector notation






Such transformations were formulated by Herglotz (1911) and Silberstein (1911) and others.

In line with equation (1b), one can substitute in (3b) or (4a), producing the Lorentz transformation of velocities (or velocity addition formula) in analogy to Beltrami coordinates of (3e):






or using trigonometric and hyperbolic identities it becomes the hyperbolic law of cosines in terms of (3f):[12][R 1][13]






and by further setting u=u′=c the relativistic aberration of light follows:[17]






The velocity addition formulas were given by Einstein (1905) and Poincaré (1905–06), the aberration formula for cos(α) by Einstein (1905), while the relations to the spherical and hyperbolic law of cosines were given by Sommerfeld (1909) and Varićak (1910).

Learning materials from Wikiversity: These formulas resemble the equations of an ellipse of eccentricity v/c, eccentric anomaly α' and true anomaly α, first geometrically formulated by Kepler (1609) and explicitly written down by Euler (1735, 1748), Lagrange (1770) and many others in relation to planetary motions.[18][19]

Lorentz transformation via conformal, spherical wave, and Laguerre transformation[edit]

If one only requires the invariance of the light cone represented by the differential equation , which is the same as asking for the most general transformation that changes spheres into spheres, the Lorentz group can be extended by adding dilations represented by the factor λ. The result is the group Con(1,p) of spacetime conformal transformations in terms of special conformal transformations and inversions producing the relation


One can switch between two representations of this group by using an imaginary sphere radius coordinate x0=iR with the interval related to conformal transformations, or by using a real radius coordinate x0=R with the interval related to spherical wave transformations in terms of contact transformations preserving circles and spheres. It turns out that Con(1,3) is isomorphic to the special orthogonal group SO(2,4), and contains the Lorentz group SO(1,3) as a subgroup by setting λ=1. More generally, Con(q,p) is isomorphic to SO(q+1,p+1) and contains SO(q,p) as subgroup.[20] This implies that Con(0,p) is isomorphic to the Lorentz group of arbitrary dimensions SO(1,p+1). Consequently, the conformal group in the plane Con(0,2) – known as the group of Möbius transformations – is isomorphic to the Lorentz group SO(1,3).[21][22] This can be seen using tetracyclical coordinates satisfying the form .

A special case of Lie's geometry of oriented spheres is the Laguerre group, transforming oriented planes and lines into each other. It's generated by the Laguerre inversion leaving invariant with R as radius, thus the Laguerre group is isomorphic to the Lorentz group.[23][24]

Learning materials from Wikiversity: Both representations of Lie sphere geometry and conformal transformations were studied by Lie (1871) and others. It was shown by Bateman & Cunningham (1909–1910), that the group Con(1,3) is the most general one leaving invariant the equations of Maxwell's electrodynamics. Tetracyclical coordinates were discussed by Pockels (1891), Klein (1893), Bôcher (1894). The relation between Con(1,3) and the Lorentz group was noted by Bateman & Cunningham (1909–1910) and others. The Laguerre inversion was introduced by Laguerre (1882) and discussed by Darboux (1887) and Smith (1900). A similar concept was studied by Scheffers (1899) in terms of contact transformations. Stephanos (1883) argued that Lie's geometry of oriented spheres in terms of contact transformations, as well as the special case of the transformations of oriented planes into each other (such as by Laguerre), provides a geometrical interpretation of Hamilton's biquaternions. The group isomorphism between the Laguerre group and Lorentz group was pointed out by Bateman (1910), Cartan (1912, 1915/55), Poincaré (1912/21) and others.

Lorentz transformation via Cayley–Hermite transformation[edit]

The general transformation (Q1) of any quadratic form into itself can also be given using arbitrary parameters based on the Cayley transform (I-T)−1·(I+T), where I is the identity matrix, T an arbitrary antisymmetric matrix, and by adding A as symmetric matrix defining the quadratic form (there is no primed A' because the coefficients are assumed to be the same on both sides):[25][26]






For instance, the choice A=diag(1,1,1) gives an orthogonal transformation which can be used to describe spatial rotations corresponding to the Euler-Rodrigues parameters [a,b,c,d] which can be interpreted as the coefficients of quaternions. Setting d=1, the equations have the form:






Learning materials from Wikiversity: After Cayley (1846) introduced transformations related to sums of positive squares, Hermite (1853/54, 1854) derived transformations for arbitrary quadratic forms, whose result was reformulated in terms of matrices (Q2) by Cayley (1855a, 1855b). The Euler-Rodrigues parameter were discovered by Euler (1771) and Rodrigues (1840).

Also the Lorentz interval and the general Lorentz transformation in any dimension can be produced by the Cayley–Hermite formalism.[R 2][R 3][27][28] For instance, Lorentz transformation (1a) with n=1 follows from (Q2) with:






This becomes Lorentz boost (4a or 4b) by setting , which is equivalent to the relation known from Loedel diagrams, thus (5a) can be interpreted as a Lorentz boost from the viewpoint of a "median frame" in which two other inertial frames are moving with equal speed in opposite directions.

Furthermore, Lorentz transformation (1a) with n=2 is given by:






or using n=3:






Learning materials from Wikiversity: The transformation of a binary quadratic form of which Lorentz transformation (5a) is a special case was given by Hermite (1854), equations containing Lorentz transformations (5a, 5b, 5c) as special cases were given by Cayley (1855), Lorentz transformation (5a) was given (up to a sign change) by Laguerre (1882), Darboux (1887), Smith (1900) in relation to Laguerre geometry, and Lorentz transformation (5b) was given by Bachmann (1869). In relativity, equations similar to (5b, 5c) were first employed by Borel (1913) to represent Lorentz transformations.

As described in equation (3d), the Lorentz interval is closely connected to the alternative form ,[29] which in terms of the Cayley–Hermite parameters is invariant under the transformation:






Learning materials from Wikiversity: This transformation was given by Cayley (1884), even though he didn't relate it to the Lorentz interval but rather to .

Lorentz transformation via Cayley–Klein parameters, Möbius and spin transformations[edit]

The previously mentioned Euler-Rodrigues parameter a,b,c,d (i.e. Cayley-Hermite parameter in equation (Q3) with d=1) are closely related to Cayley–Klein parameter α,β,γ,δ in order to connect Möbius transformations and rotations:[30]

thus (Q3) becomes:






Learning materials from Wikiversity: The Cayley-Klein parameter were introduced by Helmholtz (1866/67), Cayley (1879) and Klein (1884).

Also the Lorentz transformation can be expressed with variants of the Cayley–Klein parameters: One relates these parameters to a spin-matrix D, the spin transformations of variables (the overline denotes complex conjugate), and the Möbius transformation of . When defined in terms of isometries of hyperbolic space (hyperbolic motions), the Hermitian matrix u associated with these Möbius transformations produces an invariant determinant identical to the Lorentz interval. Therefore, these transformations were described by John Lighton Synge as being a "factory for the mass production of Lorentz transformations".[31] It also turns out that the related spin group Spin(3, 1) or special linear group SL(2, C) acts as the double cover of the Lorentz group (one Lorentz transformation corresponds to two spin transformations of different sign), while the Möbius group Con(0,2) or projective special linear group PSL(2, C) is isomorphic to both the Lorentz group and the group of isometries of hyperbolic space.

In space, the Möbius/Spin/Lorentz transformations can be written as:[32][31][33][34]