The Van Meegeren Art Forgeries Using Differential Equations:

Historical Background:

Han van Meegeren was a Dutch painter and art dealer who became one of the most ingenious art
forgers in history. Born in 1889 in Deventer, Netherlands, Van Meegeren trained in the
traditional styles of the Dutch Golden Age, but his own works received little recognition. Fueled
by bitterness and the desire to prove his talent to critics, Van Meegeren turned to forgery,
claiming his work was by famed 17th-century Dutch artists, especially Jan Vermeer. His
forgeries were so convincing that they were lauded as genuine masterpieces and acquired by
major museums and private collectors, including high-ranking Nazi officials.

During World War II, Van Meegeren sold a significant number of these forged artworks. After
the war, these sales drew the attention of the Allied authorities, who arrested Van Meegeren in
1945 on charges of collaborating with the enemy by selling Dutch cultural heritage to the Nazis.
To escape a possible death penalty, Van Meegeren made confessed that he had not sold genuine
Vermeers to the Nazis but had forged them himself. To substantiate his claims, he even painted a
new "Vermeer" while in police custody.

Scientific Investigation and Differential Equations:

The case against Van Meegeren hinged not only on legal and historical inquiries but also on
rigorous scientific analysis. Investigators needed to determine whether the paintings were
authentic Vermeers from the 17th century or modern forgeries. This is where the application of
differential equations, specifically through radiometric dating techniques, played a critical role.

Radiometric dating is based on the decay of radioactive isotopes, which is described by a first-
order linear differential equation. The decay process for a radioactive substance like lead-210,
commonly found in the white lead pigment used in historical paintings, can be modeled as:
dN
=−λN , where N(t) represents the number of radioactive atoms at time t, and λ is the decay
dt
constant, which is specific to each radioactive substance. The half-life of lead-210 is about 22
years, meaning it reduces to half its original quantity in that time. The solution to this differential
equation, given that N(0)=N0 (the initial quantity), is: N (t )=N 0 e−λt

This equation predicts how much of the radioactive material remains after a certain period.

In the case of Van Meegeren's forgeries, the investigators measured the current lead-210 levels in
the paint and compared these to the expected levels if the paintings had indeed been from the
17th century. This involved solving the decay equation to find 𝑡t, the time since the pigment
containing lead-210 was originally applied:

1
t= ln ⁡
λ ( )
N0
N (t )
This mathematical modeling showed that the amount of lead-210 was inconsistent with the
paintings being from the 17th century. This was conclusive evidence that the paintings were not
as old as claimed and were indeed forgeries. The chemical analysis of the pigments further
supported this, as modern materials were found in the compositions, which would not have been
available during Vermeer's time.

In the realm of applied mathematics, ODEs stand as main tools for modeling and understanding
dynamic systems across a broad spectrum of disciplines. These equations, which describe the
rate of change of a variable with respect to another, provide deep insights into the behavior of
complex systems, whether they involve biological growth, chemical processes, or physical
phenomena. The versatility and power of first-order differential equations make them
indispensable in both research and practical applications, where precise understanding and
prediction of system dynamics are crucial. In this context, three significant applications of first-
order differential equations: the dynamics of tumor growth, mixing problems in chemical
engineering, and the computation of orthogonal trajectories in physical systems. Each of these
applications highlights a unique aspect of how ordinary differential equations can be used to
model real-world phenomena that are fundamentally dynamic and interconnected. Those
scenarios showcase not only the mathematical utility of these equations but also their practical
implications in solving problems that impact various fields of science and engineering.

The Dynamics of Tumor Growth: This section delves into the complex nature of cancerous
tumors, whose growth patterns deviate from simple exponential growth due to biological
constraints and environmental factors. We explore how first-order differential equations model
these growth dynamics, offering insights that are crucial for medical research and treatment
strategies.

It has been observed experimentally, that "free living" dividing cells, such as bacteria cells, grow
at a rate proportional to the volume of dividing cells at that moment. Let V(t) denote the volume
dV
of dividing cells at time t. Then, =λV (1) for some positive constant λ . The solution of (1) is
dt
λ ( t −t 0)
V (t )=V 0 e (2), where V 0 is the volume of dividing cells at the initial time t0• Thus, free-
living dividing cells grow exponentially with time. One important consequence of (2) is that the
volume of cells keeps doubling every time interval of length ln2/ λ .
On the other hand, solid tumors do not grow exponentially with time. As the tumor becomes
larger, the doubling time of the total tumor volume continuously increases. Various researchers
have shown that the data for many solid tumors is fitted remarkably well, over almost a 1000-

fold increase in tumor volume, by the equation V ( t )=V 0 exp ( αλ (1−exp (−αt ) )) (3).
x
Where exp ( x ) ≡e ,∧ λ∧α are positive constants .

Equation (3) is usually referred to as a Gompertzian relation. It says that the tumor grows more
and more slowly with the passage of time, and that it ultimately approaches the limiting volume
λ/ α
V 0 e . Medical researchers have long been concerned with explaining this deviation from
simple exponential growth. A great deal of insight into this problem can be gained by finding a
differential equation satisfied by V(t). Differentiating (3) gives

dV λ
∧¿ V 0 λ exp ⁡(−αt )exp ⁡( (1−exp ⁡( −α (t))))
dt α
−αt
¿=λ e V .(4)

Two conflicting theories have been advanced for the dynamics of tumor growth. They
correspond to the two arrangements
dV
=( λ e ) V (4a)
−αt
dt
dV
=λ ( e V ) (4b)
−αt
dt
of the differential equation (4). According to the first theory, the retarding effect of tumor growth
is due to an increase in the mean generation time of the cells, without a change in the proportion
of reproducing cells. As time goes on, the reproducing cells mature, or age, and thus divide more
slowly. This theory corresponds to the bracketing (a).
The bracketing (b) suggests that the mean generation time of the dividing cells remains constant,
and the retardation of growth is due to a loss in reproductive cells in the tumor. One possible
explanation for this is that a necrotic region develops in the center of the tumor. This necrosis
appears at a critical size for a particular type of tumor, and thereafter the necrotic "core"
increases rapidly as the total tumor mass increases. According to this theory, a necrotic core
develops because in many tumors the supply of blood, and thus of oxygen and nutrients, is
almost completely confined to the surface of the tumor and a short distance beneath it. As the
tumor grows, the supply of oxygen to the central core by diffusion becomes more and more
difficult resulting in the formation of a necrotic core.
 Mixing Problems or Compartment Analysis: The application of differential equations in
scenarios where different substances are mixed, such as in a bioreactor or a chemical
processing tank. These models help in designing and optimizing processes where the control of
concentration levels is critical for product quality and safety.
Mixing problems, often encountered in engineering and biology, involve the analysis of substances
within a dynamically changing system. Consider a tank where a solution with a fixed concentration of a
substance flows in and mixes with the contents before flowing out. The concentration of the substance
dS
within the tank can be modeled by: =rate in−rate out
dt

For instance, if (𝑡) represents the amount of a substance (e.g., salt) in the tank at time t, and the
substance enters and leaves the tank at a constant rate, the concentration of the substance over
time can be determined by solving this differential equation. This is crucial for designing
systems where precise chemical concentrations must be maintained, such as in water treatment
facilities or pharmaceutical manufacturing.

Orthogonal trajectories:

In many physical applications, it is often necessary to find the orthogonal trajectories of a given
family of curves. (A curve that intersects each member of a family of curves at right angles

is called an orthogonal trajectory of the given family.) For example, a charged particle moving
under the influence of a magnetic field always travels on a curve that is perpendicular to each of
the magnetic field lines. The problem of computing orthogonal trajectories of a family of curves
can be solved in the following manner. Let the given family of curves be described by the
relation F ( x , y , c )=0 . (7)

' ' −F x
Differentiating this equation yields F x + F y y =0 ,∨ y = . (8)
Fx
Next, we solve for c=c(x,y) from (7) and replace every c in (8) by this value c(x,y). Finally, since
the slopes of curves which intersect orthogonally are negative reciprocals of each other, we see
' Fy
that the orthogonal trajectories of (7) are the solution curves of the equation y = .
Fx
 Solution Methods for ODEs:
Separation of variables:
A differential equation for the unknown f ( x ) will be separable if it can be written in the form:
d
f ( x )=g ( x ) h ( f ( x ) )
dx
where g and h are given functions. This is perhaps more transparent when written using y=f ( x )
dy dy
as: =g ( x ) h ( y ) . And since h( y)≠ 0 , we can rearrange terms to obtain: =g (x) dx , where
dx h( y )
the two variables x and y have been separated. Note dx (and dy) can be viewed, at a simple level,
as just a convenient notation, which provides a handy mnemonic aid for assisting with
manipulations.

Newton’s law of cooling. If T ( ι ) is the temperature of an object at time t and T o ul ( t ) is the

temperature of its surroundings, then
dT
=k ( T out −T )
dt

where k is a positive constant called the cooling coefficient

The Equation (1) is a first-order ODE and its general solution contains one arbitrary constant. We can see
this as follows: If Tout is a constant, then ODE (1) is separable, and separating the variables we have
dT
∫ =∫ kdt
T out −T
Finding an antiderivative for each side we obtain −ln ⁡∨T out −T (t)∨¿ kt+ K
where K is an arbitrary constant. Multiplying through by −1 and exponentiation gives us
−K −kt
¿ T out −T (t)∨¿ e e
−kt
or, after dropping the absolute value signs, we have that T ( t )=T out +C e

where C = ±e−K is now the arbitrary constant. The solution formula (2) is called the general solution of
ODE (1).
Given an initial condition, we can determine C uniquely and identify a single solution from the general
solution (2). If we take T(0) = T0, then since T ( 0 )=T out +C we see that C=T 0−T out and we get the
−kt
unique solution T ( t )=T out + ( T 0−T out ) e .
The constant of proportionality, k, in ODE (1) determines the rate at which the body cools. It can be
evaluated in a number of ways, for example, by measuring the body’s temperature at two different
times and using formula (3) to solve for T0 and k.
Exact Equations and Integrating Factors
For first-order equations, there are a number of integration methods that are applicable to various
classes of problems. The most important of these are linear equations and separable equations, which
we have discussed previously. Here, we consider a class of equations known as exact equations for
which there is also a well-defined method of solution. Keep in mind, however, that those first order
equations that can be solved by elementary integration methods are rather special; most first order
equations cannot be solved in this way.

More generally, let the differential equation M ( x , y ) + N ( x , y ) y ' =0(1) be given. Suppose that
∂ψ ∂ψ
we can identify a function ψ(x, y) such that ( x , y)=M (x , y ), (x , y)=N ( x , y ),(2)
∂x ∂y
and such that ψ ( x , y )=c defines y=ϕ ( x )implicitly as a differentiable function
of x. Then
' ∂ψ ∂ψ dy d
M (x , y )+ N (x , y) y = + = ψ [ x , ϕ( x)](3)
∂ x ∂ y dx dx
and the differential equation (1) becomes
d
ψ [ x , ϕ (x )]=0.
dx
In this case, Eq. (1) is said to be an exact differential equation. Solutions of this equation or the
equivalent, are given implicitly by ψ(x, y) = c, (9) where c is an arbitrary constant.
General Theorem:

Let the functions M , N , M y , and N x, where subscripts denote partial derivatives

be continuous in the rectangular region R :α < x< β , γ < y <δ . Then Eq.(1)
M ( x , y ) + N ( x , y ) y ' =0

is an exact differential equation in R if and only if

M y ( x , y )=N x ( x , y ) (10)
at each point of R . That is, there exists a function ψ satisfying Eqs. (2),
ψ x ( x , y )= M ( x , y ) , ψ y ( x , y )=N ( x , y ) ,

if and only if M and N satisfy Eq. (10).