Mathematical Tripos Part II Lent Term 2024

General Relativity Dr. J.M. Evans

Example Sheet 4

1. A static spacetime has line element

2
ds2 = →e2φ/c c2 dt2 + gij dxi dxj ,

where φ and gij are independent of x0 = ct, and i, j = 1, 2, 3. Show that

Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

1! ∂φ ∂φ " 1 ∂φ 2φ/c2
Γα0β = nα + nβ and Γ0i0 = g ij e ,
c2 ∂xβ ∂xα c2 ∂xj
where nα = (1, 0, 0, 0).
Let uα be the 4-velocity of a co-moving observer (i.e. an observer at rest in these coordinates, so that
u = 0 and u0 u0 = →c2 ). Show that
i

1
∇β uα = → uβ ∇α φ and so ∇α φ = uβ ∇β uα .
c2
Show further that
g αβ ∇α ∇β φ = Rαβ uα uβ
and hence that
1 ij
g ij ∇i ∇j φ + g ∇i φ∇j φ = Rαβ uα uβ .
c2
[ Hint: you may find it helpful to start from the Ricci identity uα;βγ → uα;γβ = Rδ αβγ uδ . ]
What does the last equation reduce to in the Newtonian limit (weak gravity) with Tαβ = ρ uα uβ ?

2. A perfect fluid has 4-velocity uα which is tangent to the fluid flow lines (the integral curves of uα )
and which satisfies uα uα = →c2 . If the fluid has particle number density n, density ρ and pressure p, then
the particle flux density N α and energy-momentum tensor T αβ are given by

N α = nuα , T αβ = (ρ + p/c2 )uα uβ + pg αβ ,

and both are conserved: ∇α N α = ∇β T αβ = 0.

(i) If the fluid has zero pressure, show that ∇α (ρuα ) = 0 and that the fluid flow lines are geodesics.
Show also that ρ/n is constant on each such geodesic.
(ii) If the fluid has pressure, find an expression for ∇α (ρuα ) and show that
! 1 " β 1
ρ+ p u ∇β uα + ∇α p + 2 uα uβ ∇β p = 0 .
c2 c

3. Consider a perfect fluid, with definitions and notation as in question 2, and a static, weak-field metric
of the form given in question 1, but with gij = δij . In the Newtonian limit, ϕ/c2 ∼ v 2 /c2 $ 1, where v is
a typical speed, so that uα ≈ (c, u). Show that, to lowest order,

∂n
+ ∇·(nu) = 0 ,
∂t
where ∇ is the usual vector operator in 3-dimensional flat space. What is the corresponding equation for
ρ ? Show that, in the Newtonian limit, ρ uβ ui; β = →p, i (i = 1, 2, 3) and hence that

∂u 1
+ (u · ∇)u = →∇φ → ∇p .
∂t ρ

1
 4. The Friedmann-Lemaı̂tre-Robertson-Walker (FLRW) metric with c = 1 is given by:

dr2
# $
2 2 2 2 2
ds = →dt + a(t) + r dΩ
1 → kr2

and
3(ȧ2 + k) 2aä + ȧ2 + k
Gtt = , Grr = → .
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

a2 1 → kr2
For a dust universe with Ttt = ρ, show that ρ a3 = ρ0 , where ρ0 is a constant.
(i) In the case k = 0, show that aȧ2 = A2 , where A is a constant, and deduce that the universe expands
for ever. Without further calculation, explain how this conclusion is affected in the case k < 0.
dη 1
(ii) In the case k > 0, we define a new time coordinate η by = , where R2 = k −1 . Derive the
dt Ra
equations
a(η) = B(1 → cos η) , t(η) = BR(η → sin η) ,
where B is a constant, and hence show that the universe recollapses within a finite time.
(iii) For the solution in (ii), set r = R sin ϕ in the line element and use the formula for the 3-space volume
element
dV = (gχχ gθθ gφφ )1/2 dϕ dθ dφ
to determine the volume of the universe at a given scale factor (the angular coordinates run from 0 to π
for ϕ and θ, and from 0 to 2π for φ). Hence find the maximum volume in terms of M G, where M is the
total mass of the universe, and use dimensional analysis to restore the dependence of the result on c.

5. Obtain the geodesic equations for the closed (k = 1) FLRW dust universe, using η, ϕ, θ, φ coordinates
and show that there are null geodesics with θ = ϕ = 12 π. How many times can a photon encircle the
universe from the time of creation to the moment of annihilation?

6. Show that the Einstein-Maxwell equations (i.e. the Einstein equations with energy momentum tensor
for an electromagnetic field T αβ = F αγ F β γ → 14 F γδ Fγδ g αβ ) can be written

Rαβ = κ Fαγ Fβ γ → 41 gαβ Fγδ F γδ .

% &

For a line element of the form

1
ds2 = →f (r)c2 dt2 + dr2 + r2 (dθ2 + sin2 θ dφ2 ) ,
f (r)

the only non-zero components of the Ricci tensor are given by

Rtt /(c2 f ) = →f Rrr = 12 f "" + f " /r , Rθθ = Rφφ / sin2 θ = 1 → rf " → f .

In the case
Q
Ftr = →Frt = and Fαβ = 0 otherwise,
r2
show that a solution can be found that reduces to the Schwarzschild solution when Q = 0.
Find an analogous solution in the case Rαβ = Λgαβ .

2
 7. For the Schwarzschild metric, a retarded time coordinate u is defined by u = ct → r∗ , where
dr/dr∗ = F (r) = 1 → 2M/r. Show that, with this definition, the line element can be written

ds2 = →F du2 → 2du dr + r2 (dθ2 + sin2 θ dφ2 ) .

Consider a spacecraft that is freely falling radially into a Schwarzschild black hole, with 4-velocity V α
and proper time τ . The spacecraft emits monochromatic radio signals, of wavelength λe , which propagate
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

radially outwards and are received, with wavelength λo , by a distant observer who is at rest with respect
to the Schwarzschild coordinates.
Show that
λo ∆to ∆uo ∆ue Vu
= = = ≈
λe ∆τ c∆τ c∆τ c
where, for example, ∆to is the proper time interval during which the observer receives one cycle of the
signal and ∆τ is the time for the spacecraft to emit one cycle.
Now show that Vu = →K, where K is a constant, and that
√
K + K 2 → F c2 '
Vu = , V r = → K 2 → F c2 .
F

Deduce that on the world line of the spacecraft near the horizon du/dr ∼ →2/F , and that u ∼ →2r∗
and F ∼ e−u/(4M ) .
Conclude that, just as the spacecraft is about to cross the event horizon, the observer sees the frequency
red-shifted with an observer-time dependence ∝ exp(→ct/(4M )).

8. Show that, for an observer with proper time τ moving in the Schwarzschild spacetime,

c2 = F c2 ṫ2 → ṙ2 /F → r2 (θ̇2 + sin2 θ φ̇2 ),

where ṫ = dt/dτ etc., and F = 1 → 2M/r. Show, that for an observer within the Schwarzschild horizon,
ṙ2 ≥ →c2 F however the observer moves. Deduce that any observer crossing the Schwarzschild horizon will
reach r = 0 within a proper time πM/c.

9. Let M be the torus (S 1 × S 1 ) and define the metric gαβ on M by

ds2 = sin θ (dφ2 → dθ2 ) + 2 cos θ dθ dφ ,

where 0 ≤ θ ≤ 2π and 0 ≤ φ ≤ 2π. Show that, for a null geodesic,

φ̇2 + 2 φ̇ θ̇ cot θ → θ̇2 = 0 ,

where dot is differentiation with respect to an affine parameter, and deduce that the curves given by
φ = →2 ln sin(θ/2) + φ0 and φ = →2 ln cos(θ/2) + φ0 are null geodesics. Use another first integral of the
Euler-Lagrange equations to show that in both cases θ = pλ, where λ is an affine parameter and p is a
constant.
Show that one family of null geodesics wraps round the torus an infinite number of times within a
finite range of the affine parameter, never reaching the null curve θ = 2π, and that the other family of
null geodesics crosses this curve.
Is this space geodesically complete? Is the Riemann tensor well-behaved (no calculation required)?

3
 10. (i) A weak gravitational field has the spacetime metric gαβ = ηαβ + -hαβ + O(-2 ), where ηαβ is the
Minkowski metric and - is small. Show that

Rαβγδ = 1
2 -( hαδ,βγ + hβγ,αδ → hαγ,βδ → hβδ,αγ ) + O(-2 ) .
γ
Let h = hγ γ and define hαβ = hαβ → 12 h ηαβ . Check that hαβ = hαβ → 21 h ηαβ where h = h γ, and show
that
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

Rαβ = 12 -( →! hαβ + hα γ ,βγ + hβ γ ,αγ + 12 ηαβ ! h ) + O(-2 ) ,

where ! = η αβ ∂α ∂β . What is the linearised vacuum Einstein equation for hαβ ?
(ii) An infinitesimal coordinate transformation, which is also known as a gauge transformation, is given
by xα → xα → -f α (x). Show that

hαβ → hαβ + fα,β + fβ,α + O(-) ,

but that the curvature tensors are unchanged to leading order in -. Deduce that if f α is chosen to satisfy
!f α = →hαβ ,β , then in the new coordinates the gauge condition hαβ ,β = 0 holds. Conclude that, with
this choice, the linearised vacuum Einstein equation for weak fields is the wave equation:

! hαβ = 0 .

β
(iii) Consider a gravitational wave solution hαβ = Hαβ eikβ x with Hαβ,γ = 0 (note: this is an ansatz
for hαβ , not hαβ ). Show that, in order to satisfy both the linearised Einstein equation and the gauge
condition in (ii), k α must be a null vector and Hαβ k β = 12 kα Hβ β must hold.
(iv) Corresponding to the remaining freedom to make gauge transformations, show that there is an arbi-
trariness in the solution given by Hαβ → Hαβ + kα vβ + vα kβ for any vα . How many degrees of freedom
are there for a gravitational wave propagating in a given direction k α ? If k α = k(1, 0, 0, 1), show that we
may take the independent components of Hαβ to be H11 = →H22 and H12 = H21 .

Comments to: [email protected]

4
 Mathematical Tripos Part II Lent Term 2024
General Relativity Dr. J.M. Evans

Example Sheet 2

1. (i) Consider a parametrised curve given by xµ (λ) or x̃α (λ), depending on the choice of coordinates.
Check that T µ = dxµ /dλ and T̃ α = dx̃α /dλ are related by the transformation rule for components of a
vector.
Using, in addition, the relation between the metric components gµν and g̃αβ in the coordinates above,
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

check that Tµ = gµν T ν and T̃α = g̃αβ T̃ β are related by the transformation rule for components of a
covector.
(ii) With notation as (i), start from the geodesic equation in coordinates xµ ,

dT µ
+ Γνµρ T ν T ρ = 0 ,
dλ

substitute for T µ in terms of T̃ α , then compare with the geodesic equation in coordinates x̃α to show that

∂xµ α ∂xν ∂xρ µ ∂ 2 xµ

Γ̃ = Γ + .
∂ x̃α β γ ∂ x̃β ∂ x̃γ ν ρ ∂ x̃β ∂ x̃γ

2. Let xµ be coordinates on a manifold of dimension n, and let gµν be components of a metric with a
Taylor expansion about the origin:
1 !
gµν = ηµν + Cµνρ xρ + C xρ xσ + . . . .
2! µνρσ
Consider a change of coordinates, also given by a Taylor expansion:
1 α 1 !α
x̃α = xα + A βγ xβ xγ + A βγδ xβ xγ xδ + . . . .
2! 3!
What can be assumed about symmetries of the coefficients C, C ! , A and A! , without loss of generality?
With these assumptions, show that C and A each have 21 n2 (n + 1) independent components. Find
1 2 2
the number of independent components of C ! and A! and hence show that the difference is 12 n (n → 1).
What can be deduced from these observations?

3. A static spacetime has line element

2
ds2 = →e2φ/c c2 dt2 + gij dxi dxj

where φ and gij are functions only of xi , with i, j = 1, 2, 3, while x0 = ct. Show that

1 ∂φ
Γ00i =
c2 ∂xi
and find Γ000 , Γ0i0 and Γi0j , where Γβαγ is the Levi-Civita connection (give your answer in terms of g ij , the
inverse of gij , where necessary).
An observer is at rest in the coordinate system above and has 4-velocity V α = dxα /dτ , where τ is
proper time. Use the normalisation condition V α Vα = →c2 to find V 0 and V0 .
The 4-acceleration is defined by Aα = (dV α /dτ ) + Γβαγ V β V γ . Show that Aα = ∂φ/∂xα .

1
 4. If Aα is a covector field, let
∂Aβ ∂Aα
Fαβ = → .
∂xα ∂xβ
By writing this expression in terms of covariant derivatives, or otherwise, show that Fαβ is a tensor field.
If U α and V α are vector fields, the commutator [ U, V ]α is defined by

∂V α ∂U α
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

[ U, V ]α = U β β
→ Vβ .
∂x ∂xβ
Show that [ U, V ]α is a vector field.

5. The metric gαβ (x) has the property that, if each point xα is mapped to φα (x), distances are unaltered.
Such a mapping is called an isometry for this metric. Show that

! " ∂φα ∂φβ

gαβ φ(x) = gγδ (x) .
∂xγ ∂xδ
Setting φα (x) = xα + &ξ α (x), where & is small, show that to first order in &

∂gαβ ∂ξ γ ∂ξ γ
ξγ γ
+ gγβ α + gγα β = 0 .
∂x ∂x ∂x
Show also, using the formula for the connection components, that this condition can be written in tensorial
form as
∇α ξβ + ∇β ξα = 0 .
This is known as Killing’s equation and solutions ξα are called Killing covector fields.
(i) Consider a massive, freely falling particle moving along an affinely parametrized geodesic with
tangent vector V α . Show that Killing’s equation implies that E = ξα V α is constant along the geodesic.
(ii) If a Killing vector field takes the form ξ α = (1, 0) in a given coordinate system, show that
∂gαβ /∂x0 = 0. What is the physical interpretation of E in this case?

6. Write down the radial Euler-Lagrange equation for the Schwarzschild metric and show that, in the
case of a circular geodesic orbit in the equatorial plane, this determines the period of the orbit in terms
of Schwarzschild coordinate time. How does this relate to the Newtonian result?
Alice orbits the Earth on a circular geodesic path of radius R, while Bob stays at home on the Earth’s
surface, at radius RE (the Earth is taken to be a non-rotating sphere).
(i) Show that Alice’s proper time τA is related to the Schwarzschild time coordinate t by

dτA2 = (1 → 3M/R) dt2 .

(ii) How is Bob’s proper time τB related to the coordinate t ? Deduce that Bob and Alice age at the
same rate if R = 32 RE .

2
 7. A massive particle moves on a circular orbit of radius R in Schwarzschild spacetime, with E and
h denoting the usual first integrals for t and φ (take units with G = c = 1). Use the r Euler-Lagrange
equation to find conditions that R must satisfy, and show that they are the same as would be obtained
by considering motion in the 1-dimensional potential Veff (r), where
# h2 $# 2M $
2Veff (r) = 1+ 1 → → 1.
r2 r
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

(i) Express h in terms of R and M and deduce that there are circular orbits for R > 3M .
(ii) Show that these orbits are stable if R > 6M and unstable if 3M < R < 6M .
(iii) Show that the fractional binding energy (i.e. 1→E) of a stable circular orbit in the limit R → 6M is
√
(1 → 2 2/3) % 0.0572.

8. A massive test particle flies past a spherically symmetric star of mass M and Schwarzschild radius
rs = 2M (using units in which G = c = 1), which causes a small deflection in the particle’s trajectory.
Using the radial equation with effective potential Veff (r) and setting u = 1/r, show that
# du $2 E2 → 1 2
# 1
2
$
= → u + rs u + u .
dφ h2 h2
%∞
Now consider a solution of the form u(φ) = n=0 &n un (φ), where & & 1.

(i) If rs u & 1, the form of the trajectory to leading order can be found by setting rs = 2M = 0. Show
that the resulting solution is u0 (φ) = (1/b) sin φ, with a suitable choice of the coordinate
√ φ, where b
is a constant. Show also that if the speed of the particle is v when r = b, then h = bv/ 1 → v 2 .
(ii) Setting & = rs /b & 1 and taking u0 (φ) as in (i), find a second order differential equation for u1 (φ).
Hence show that, to order &, the angular deflection is given by
# b2 $ rs (1 + v 2 )
∆φ ≃ 2& 1 + 2 = .
2h bv 2

Comment on the limit v → 1.

9. In an attempt to unify Special Relativity and Newtonian Gravity, the orbits in a central potential are
calculated using the Euler-Lagrange equations derived from the Lagrangian
GM
L = →c2 γ −1 +
r
where γ = (1 → v·v/c2 )−1/2 (in the obvious notation).
Obtain the orbital equation for motion in the equatorial plane in the form

u!! + u = γ )−1

where u = 1/r, prime denotes differentiation with respect to φ, the usual angular coordinate in the plane,
) = h2 /GM , and h is a conserved quantity, to be specified.
"1/2
Show that γ = 1 + h2 (u!2 + u2 )/c2 . Using the approximation γ ≃ 1 + 12 h2 (u!2 + u2 )/c2 , show
!

that the rate of advance of the perihelion is one sixth of that obtained using the Schwarzschild metric.

3
 10. A classical test of general relativity is the time delay caused to radar signals bounced off planets
or satellites. Ignoring curvature a light ray is given by r sin φ = b. By differentiating this, show that
r2 dφ2 = b2 dr2 /(r2 → b2 ).
Assuming the above relation between dφ and dr obtain the approximate equation for light rays in
Schwarzschild
M b2
& '
r 2M
dt = ± √ 1+ → 3 dr ,
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

r 2 → b2 r r
where O (M/r)2 terms have been neglected. Show that, to this approximation, the time taken to move
! "

from r = b to r = r1 is given by
M
( (
∆t = r12 → b2 + 2M cosh−1 (r1 /b) → r12 → b2 ,
r1
and identify the first term on the right hand side.

Comments to: [email protected]

4
 Mathematical Tripos Part II Lent Term 2024
General Relativity Dr. J.M. Evans

Example Sheet 3

1. The Ricci identity for a vector field V µ is →α →β V µ − →β →α V µ = Rµ ναβ V ν .

(i) Deduce the corresponding result for a covector field Wµ by considering →α →β (V µ Wµ ) − →β →α (V µ Wµ ). Is there
an easier way to find this result?
(ii) Given two vector fields U µ and V µ , evaluate →α →β (U µ V ν ) − →β →α (U µ V ν ) . Deduce that, for an arbitrary
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

tensor field T µν ,
→α →β T µν − →β →α T µν = Rµ σαβ T σν + Rν σαβ T µσ .
Hence show that →α →β T αβ = →β →α T αβ , for any tensor field T αβ .

2. Show that, if a vector S α is parallelly transported along an affinely parametrized geodesic xα (λ) with tangent
vector T α , then gαβ S α T β is constant along the curve.
Consider the parallel transport of a vector S α = (S θ , S φ ) around a closed path on the unit 2-sphere, with θ, φ the
usual polar coordinates. The path consists of the following four segments:
(i) θ = 21 π, φ0 ≥ φ ≥ 0 , (ii) 12 π ≥ θ ≥ θ0 , φ = 0 , (iii) θ = θ0 , 0 ≤ φ ≤ φ0 , (iv) θ0 ≤ θ ≤ 21 π, φ = φ0 ,
and S α = (1, 0) at the starting point, θ = 21 π, φ = φ0 (on the equator).
(a) Sketch a picture in the case θ0 = 0 (so the path is a spherical triangle with one vertex at the North pole) using
the result of the first paragraph (no further calculation required) and hence show that the angle between the initial
and final vectors S α is proportional to the area enclosed by the path.
(b) Verify that for 0 < θ0 < 12 π the parallel transport equation has the following solutions for S α on each segment:
(i) (1, 0) , (ii) (1, 0) , (iii) ( cos(c0 φ) , − sin(c0 φ)/ sin θ0 ) , (iv) ( cos(c0 φ0 ) , − sin(c0 φ0 )/ sin θ ) , where c0 = cos θ0 .
Write down S α at the end point of the path and check that, when θ0 → 0, this agrees with the result in part (a).
[ The non-zero connection components on the 2-sphere are Γφθφ = − sin θ cos θ and Γθφφ = Γφφθ = cot θ . ]

3. Show, by considering its symmetries, that the Riemann curvature tensor for a metric on a 2-dimensional manifold
has only one independent component. Show further that for such a metric
1
Rαβγδ = R(gαγ gβδ − gαδ gβγ ) .
2

Verify this result using the connection components for 2-dimensional de Sitter spacetime (obtained in question 9 on
Example Sheet 1).

4. Let φ be a scalar field in curved spacetime such that

→α →β φ = Rαβ ,

where Rαβ is the Ricci tensor. Show that

→α (→β →β φ) = −2Rαβ →β φ

and hence deduce that →α φ→α φ + R is constant.

[ Hint: use the Ricci identity and the contracted Bianchi identity, →β Rαβ = 12 →α R . ]

1
 5. The Maxwell tensor Fαβ for the electromagnetic field in curved spacetime is given in terms of a vector potential
Aα by Fαβ = →α Aβ − →β Aα . Show that this implies →[γ Fαβ] = 0. Show further that if the energy momentum tensor

1
T αβ = F α γ F βγ − g αβ Fγδ F γδ ,
4
is conserved, i.e. →β T αβ = 0, then this implies →β F αβ = 0, provided that the matrix F α β is non-singular.
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

6. Let ξα be a Killing covector field, satisfying ξα;β + ξβ;α = 0 (see question 5 on Example Sheet 2). Use the Ricci
identity and Rα [βγδ] = 0 to show that
ξα;βγ = −Rδ γαβ ξδ .
In the case of Minkowski space, integrate this equation twice and deduce that there are 10 independent Killing vectors.

7. Use local inertial coordinates to prove that

Rαβγδ = Rγδαβ .

8. Consider Newtonian spacetime with (Cartesian) coordinates xα = (t, xi ). Show that the motion of a freely-falling
particle can be described by a curve xα (λ), where

d2 t d2 x i ∂Φ ! dt "2
= 0, + = 0,
dλ2 dλ 2 ∂xi dλ
for a suitable parameter λ and Newtonian gravitational potential Φ(xi ). Regarding this as a geodesic equation, read
off the Newtonian connection components and deduce that the corresponding curvature is given by

∂2Φ
Ri 0j0 = −Ri 00j = , Rα βγδ = 0 otherwise .
∂xi ∂xj
Can this Newtonian connection and curvature arise from a metric? [ Hint: For the standard Levi-Civita connection,
what symmetries does the Riemann tensor possess? ]

9. The Lie derivative (Lξ V )α of a vector field V α with respect to a vector field ξ α (assumed to be timelike) is defined
by the following conditions:
(i) If {xα } is a coordinate system in which ξ α = (1, 0, 0, 0), then

∂V α β ∂V
α
(Lξ V )α = = ξ ,
∂x0 ∂xβ
and (ii) (Lξ V )α transforms as a vector. Show that, in a general coordinate system,

(Lξ V )α = ξ β →β V α − V β →β ξ α .

Suppose, in addition, that the Lie derivative Lξ φ of a scalar field φ with respect to a vector field ξ α is defined in a
general coordinate system {xα } by
∂φ
Lξ φ = ξ α α
∂x
and that the Lie derivative obeys the usual Leibniz rule when applied to a tensor product. Find the Lie derivative
(Lξ U )α of a covector field Uα .
Write down an expression for the Lie derivative with respect to ξ α of a (0, 2) tensor Tαβ and show that the condition
for ξ α to be a Killing vector field (as in question 6 above) is (Lξ g)αβ = 0, where gαβ is the metric tensor.

2
 10. (i) Let M be an invertible matrix. Show that under a small change δM , the corresponding change in the
determinant is, to first order, δ(detM ) = (detM ) tr(M −1 δM ) .
[ Hint: if the entries of a matrix A are small then, to first order, det(I + A) = 1 + trA, where I is the identity matrix. ]
(ii) Let gαβ be a metric with Lorentzian signature and let g = det(gαβ ). Use the result in (i) to show that

1 ∂g 1 ∂ √
Γαββ = α
= √ −g ,
2g ∂x −g ∂xα
Copyright © 2023 University of Cambridge. Not to be quoted or reproduced without permission.

where Γαγβ is the Levi-Civita connection. (Note that g < 0 for a metric with Lorentzian signature.)
(iii) A tensor density of weight q is defined to be a quantity that transforms as a tensor under a change in coordinates
from {xµ } to {x̃α } but with an additional factor of ∆q , where ∆ = det(∂xµ /∂ x̃α ) , the Jacobian. Show that g transforms
as a scalar density of weight 2.
(iv) For ψ a scalar density of weight q, the covariant derivative is defined by

∂ψ
→α ψ = − q Γαββ ψ .
∂xα
Show that →α ψ is a covector density of weight q.

Comments to: [email protected]