On the deflection of light in a gravitational field
Samir Askerov
March 2025
1 Introduction
In this paper, I aim to demonstrate how weak gravitational effects in General Relativity can be approximated
by a theory in which the velocity of light varies near masses. To do this, I will utilize concepts derived from
a previous publication, specifically the scalar field that represents a refractive index in space induced by
gravitational fields. I will show how this theory can be applied to solve a well-known problem: the deflection
of light near the Sun. Although I have previously calculated this phenomenon, the result was not entirely
convincing and relied on certain assumptions. Therefore, I will strive to make this derivation as fundamental
as possible, relying solely on the postulate that gravitational fields create a refractive index in space. While
it may seem that such an alternative theory for relativistic effects is redundant or unproductive, I aim to
demonstrate that it offers the advantage of being significantly simpler in terms of mathematical complexity
for weak fields, when compared to the more established theory of General Relativity.
2 Light trajectories and Fermat’s principle
Recalling our previous findings, we know that we can assume that every gravitational potential is accompanied
by a scalar field η(x, y, z), representing the refractive index. It is fair to assume light rays travel along one
plane, and as such η = η(x, y), where the latter is given by
α
η(x, y) ≈ 1 + (1)
r(x, y)
p
where α = 2GM c20
is the Schwarzschild radius of the mass and r(x, y) = x2 + y 2 . With this in hand, we
can proceed with the description of our system. We imagine a light ray traveling along a path y(x) on the
xy plane. This ray will arrive from −∞ with velocity c0 , and approach the sun (point-like), of mass Msun .
The light ray will then be deflected by the gravitational field and propagate off to +∞ along a straight
line. Let b represent the impact parameter, meaning that the light ray is closest to the sun at distance
b ≈ Rsun . According to our postulate, the electromagnetic wave slows down near the sun by a factor of η1
c0
which means that the velocity along the trajectory S is given by dS dt = η(x,y) . By rearranging and integrating
this derivative in the following manner
η(x, y)
dt = dS
c0
Z t Z
1
⇒ dt = η(x, y)dS
t0 c0 S
Z
1 p
= η(x, y) dx2 + dy 2
c0 S
s
1 +∞ dy 2
Z
= η(x, y) 1 + dx
c0 −∞ dx
we obtain the following expression
Z +∞
1 p
τ= η(x, y) 1 + y ′2 dx (2)
c0 −∞
, characterising the time τ it took the light ray to travel along its trajectory y.
1
According to Fermat’s principle: ”light travels between two points along the path that requires the least
time, as compared to other nearby paths”. In our case, the former statement implies that the path y which
guides the light, is such that τ is minimized. In mathematical terms this means that δτ = 0. According to
equation 2 we have that Z +∞ p
δ η(x, y) 1 + y ′2 dx = 0 (3)
−∞
p
From equation 3, we clearly see that the function y(x) can be found by defining L(x, y, y ′ ) := η(x, y) 1 + y ′2
as our Lagrangian and by applying the Eurler-Lagrange equation to L. Now we can find the differential
equation governing the motion of our light ray.
3 Applying the Euler-Lagrange equation
In the previous section, we were able to define our Lagrangian L(x, y, y ′ ). According to the Euler-Lagrange
equations, we have
d ∂L ∂L
=
dx ∂y ′ ∂y
The right can be found to be
∂L ∂η ∂r p
= 1 + y ′2
∂y ∂r ∂y
αp
= −y 3 1 + y ′2 (4)
r
In similar fashion, the left side can be found
∂L y′
′
= ηp
∂y 1 + y ′2
!
y′ y′
d ∂L dη d
⇒ = +η
dx ∂y ′
p p
dx 1 + y ′2 dx 1 + y ′2
!
y ′′ y ′2 y ′′
∂η ∂r ′ ∂r ′2 1
= y + y +η − (5)
(1 + y ′2 )3/2
p p
∂r ∂x ∂y 1 + y ′2 1 + y ′2
Equating (4) and (5) together yields the following differential equation
!
y ′′ y ′2 y ′′
∂η ∂r p ∂η ∂r ′ ∂r ′2 1
1 + y ′2 = y + y +η p −
(1 + y ′2 )3/2
p
∂r ∂y ∂r ∂x ∂y 1 + y ′2 1 + y ′2
1
⇔ −αxy ′ + αy + ηr3 y ′′ =0 (6)
1 + y ′2
where we have assumed that r > 0. We can now solve this differential equation for y.
4 Solving the ODE
