Why a Meeting Point Matters
A direction field assigns a slope to each point in the plane. A solution curve follows those assigned slopes. If two solution curves meet at one point, then at that point they share both the same \(x\)-coordinate and the same \(y\)-value. They therefore satisfy the same initial condition there.
That observation connects the geometry of the curves to a central question about differential equations: does the initial condition determine only one solution? If it does, two solutions through the same point cannot separate after meeting. The conclusion depends on uniqueness actually holding in the region under consideration; it is not a rule that applies to every differential equation.
Initial-Value Problems and Uniqueness
A first-order differential equation in explicit form is written \(y'=f(x,y)\). A solution on an interval \(I\) is a differentiable function \(y:I\to\mathbb{R}\) that satisfies the equation at every \(x\in I\). An initial-value problem adds a prescribed value \(y(x_0)=y_0\), where \(x_0\in I\).
Uniqueness means there is at most one solution for the initial condition; existence is a separate issue. There might be no solution, exactly one solution, or more than one solution. The claim that curves cannot cross under uniqueness concerns the last possibility: uniqueness rules out two different solutions through the same initial point.
A standard sufficient condition for local uniqueness is that \(f\) be continuous and locally Lipschitz in its second variable near \((x_0,y_0)\). Locally Lipschitz in \(y\) means that in a sufficiently small neighborhood there is a constant \(L\geq 0\) such that, for the relevant \(x\) and \(u,v\),
This condition, together with continuity, is part of the standard existence and uniqueness theorem for first-order initial-value problems. It is a sufficient condition, not a necessary one: uniqueness can hold even when this particular Lipschitz condition does not.
The Noncrossing Result
The proof is short, but it expresses the essential logic behind the geometric claim.
At \(x=x_0\), both solutions have value \(y_0\), so both satisfy \(y'=f(x,y)\) together with \(y(x_0)=y_0\).
By the assumption that this initial-value problem has uniqueness on \(I\), any two solutions satisfying it are equal throughout \(I\).
Since \(y_1(x)=y_2(x)\) for every \(x\in I\), they cannot meet at \(x_0\) and then differ elsewhere on that interval.
In particular, if two solution curves meet at some point \((x_0,y_0)\) in a region where the uniqueness theorem applies, that point can be used as the initial point. The two curves must agree at least where the uniqueness conclusion applies. If uniqueness is guaranteed throughout their common interval, they agree throughout that interval.
This is a statement about solutions of the same differential equation in a region where the required uniqueness result holds. It does not claim that curves can never intersect outside that region, or that every equation has uniqueness at every point.
Worked Examples Under Uniqueness
Worked Example: Comparing Two Solutions of \(y'=y\)
Consider \(y'=y\), with the initial condition \(y(0)=1\). The function \(y_1(x)=e^x\) satisfies the equation because
It also satisfies \(y_1(0)=e^0=1\). The function \(f(x,y)=y\) is continuous, and it is locally Lipschitz in \(y\), since
Now suppose another solution \(y_2\) satisfies the same initial condition. Both solutions solve the same initial-value problem, so uniqueness gives \(y_2(x)=y_1(x)=e^x\) wherever the uniqueness conclusion applies. For a direct check that the explicit solutions \(e^x\) and \(2e^x\) do not cross, solve the proposed intersection equation:
which is impossible because \(e^x\neq 0\). The initial values are different as well: \(e^0=1\), while \(2e^0=2\). Thus these two solutions do not share an initial point at \(x=0\), and their failure to intersect is consistent with the uniqueness conclusion.
Worked Example: Intersecting Members of a Solution Family
Consider the equation \(y'=2x\). For any constant \(C\), the function \(y_C(x)=x^2+C\) is a solution. Indeed,
Take two members, \(y_1(x)=x^2+1\) and \(y_2(x)=x^2-3\). If they crossed, their values would be equal at some \(x_0\). But equality would require
which is impossible. More generally, if two members \(x^2+C_1\) and \(x^2+C_2\) have the same value at \(x_0\), then
Therefore the two functions are identical, not distinct curves crossing at a point. The equation’s right-hand side \(f(x,y)=2x\) is continuous and locally Lipschitz in \(y\), since \(f(x,u)-f(x,v)=0\). The computed solution family agrees with the uniqueness conclusion.
When Curves Can Cross
The uniqueness hypothesis matters. Consider the equation \(y'=3y^{2/3}\), where \(y^{2/3}\) is interpreted as \((\sqrt[3]{y})^2\), so it is defined for every real \(y\). Two functions are \(y_1(x)=0\) and \(y_2(x)=x^3\). For the first,
For the second,
Thus both functions solve the equation. They both satisfy the initial condition \(y(0)=0\), but they are not the same function: for example, \(y_1(1)=0\) while \(y_2(1)=1\). Their graphs meet at \((0,0)\) and differ on either side of it. Here the initial-value problem is not unique.
The right-hand side \(f(x,y)=3y^{2/3}\) is continuous, but it is not locally Lipschitz in \(y\) near \(0\). If a Lipschitz constant \(L\) worked near \(0\), taking \(u=h>0\) and \(v=0\) would require
For any fixed finite \(L\), the right-hand side \(Lh^{1/3}\) becomes smaller than \(3\) when \(h>0\) is sufficiently small. The required inequality therefore cannot hold throughout a neighborhood of \(0\). This example shows why continuity alone does not ensure uniqueness.
Worked Example: Checking a Claimed Crossing
Suppose someone claims that \(y=0\) and \(y=x^3\) cannot meet because solutions of a first-order equation never cross. Check the claim for \(y'=3y^{2/3}\).
First, at \(x=0\), both functions have value \(0\), so they share the initial condition \(y(0)=0\). Next, differentiate and substitute. For \(y=0\), both sides are \(0\). For \(y=x^3\), the derivative is \(3x^2\), and the right-hand side is \(3(x^3)^{2/3}=3x^2\). Both are solutions, but \(0\neq x^3\) when \(x\neq0\).
The claimed rule is too broad. The correct conclusion is conditional: curves sharing an initial point must agree when the initial-value problem has uniqueness in the region where the conclusion is being used. In this example uniqueness fails at the shared initial point, so the crossing does not contradict the theorem.
A Common Mistake: Treating Noncrossing as Unconditional
A tempting but incorrect argument is: “A differential equation gives one slope at each point, so only one curve can pass through that point.” A specified slope fixes the derivative at that point, not necessarily the entire curve. In the example \(y'=3y^{2/3}\), both \(y=0\) and \(y=x^3\) have slope \(0\) at the origin and satisfy the equation, yet they are different solutions with the same initial value.
The correction is to check the uniqueness hypothesis before drawing the conclusion. A single direction at a point does not by itself prove there is only one solution through that point. Under conditions that guarantee uniqueness, the theorem rules out distinct solutions through it; without those conditions, further analysis is needed.
How to Use the Result
When studying a direction field or a collection of solution curves, use the following reasoning:
- Identify the differential equation and the point where two proposed curves meet.
- Check that both curves satisfy the equation in the region being considered.
- Recognize that their shared point gives them the same initial condition.
- Determine whether uniqueness is guaranteed there, for example by the standard local Lipschitz condition.
- If uniqueness applies, conclude that the curves agree wherever the uniqueness conclusion covers them. If it does not, do not assume either that they must cross or that they cannot.
Noncrossing is useful because it lets us interpret a direction field consistently: under uniqueness, a solution curve through a given point is determined locally, so a second distinct solution cannot arrive at that same point and depart along a different path. This reasoning applies only within the domain and interval covered by the uniqueness result. A singularity, a boundary of the equation’s domain, or failure of the uniqueness condition can change what conclusions are justified.
Check Your Understanding
Answer each question using the uniqueness condition, not just the visual appearance of a curve.
- What initial-value problem do two solution curves satisfy if they meet at \((x_0,y_0)\)?
- State a standard sufficient condition on \(f\) that guarantees local uniqueness for \(y'=f(x,y)\).
- If uniqueness holds for the initial-value problem at a shared point, what does the theorem say about the two solutions on the interval covered by that uniqueness conclusion?
- For \(y'=3y^{2/3}\), verify which shared initial condition is satisfied by \(y=0\) and \(y=x^3\), and explain why their crossing does not violate the noncrossing theorem.
- Why does knowing that the equation assigns a single slope at each point not, by itself, prove that the initial-value problem has a unique solution?
