Notes 4 – More on planes and lines.

In this lecture we investigate how lines and planes relate to each other. We will also see how
to deal with symmetry and projection with respect to planes and lines.

L4.1 Parallelism and orthogonality. We presented a plane as the set of solution of one linear
equation
π : ax + by + cz = d
→
− →
− →
−
and we introduced the normal vector of the plane → −n = a i + b j + c k . In contrast with
this situation, a line r in three space is the solution set of a linear system having two linearly
independent equations, 
a1 x + b1 y + c1 z = d1
r: ;
a2 x + b2 y + c2 z = d2
thinking that a line in three space is presented by one equation is one of the worst thing you
can do in this class! A parametric equation of this line can be derived using the direction
vector →
−p =− →1 × −
n →2 , thus
n
→
−
v =→ −
v0 + t→
−
p;
as t varies the position vector →−
v describes the lines r .
By the geometric meaning of the normal vector and the direction vector, it is easy to see that
(i) The planes α, β are parallel if and only if −n→ −
→
α knβ .
(ii) The planes α, β are orthogonal if and only if − n→ −→
α ⊥ nβ .
(iii) The plane α and the line r are parallel if and only if − n→ →
−
α ⊥ vr .
(iii) The plane α and the line r are orthogonal if and only if nα k→ −
→ −vr .
→
− →
−
(iv) The lines r, s are parallel if and only if vr k vs .
(v) The lines r, s are orthogonal if and only if → −vr ⊥ → −
vs .

L4.2 Intersection of planes and lines. As lines and planes are defined as solution sets of
linear equations, in order to intersect them it is enough to solve a (larger) linear system of
equation. For the time being, we will solve these systems of equations by hands, but a more
systematic way of studying them will be soon introduced.
The intersection of two planes α and β is computed by solving a 2 × 3 linear system of
equations, that is a system with 2 equations in 3 unknowns. If the normal vectors − n→
α and
−
→
nβ are proportional, then the planes are either parallel, and we will find no solution, or they
coincide, and we will find infinitely many solutions depending on two parameters. If the
normal vectors are not parallel then α ∩ β is a line.
The intersection of a plane α and a line r is obtained by solving a 3 × 3 linear system of
equations, that is a system with 3 equations in 3 unknowns. If the normal vector of the
plane → −
n and the direction vector of the line → −
p are orthogonal, then either the plane and
the line are parallel, and we will find no solution, or the line lies inside the plane, and we
will find infinitely many solutions depending on one parameter. If the normal vector of the
plane →−n and the direction vector of the line →
−
v are not orthogonal, α ∩ r is exactly one point
and the system has exactly one solution.
The intersection of two lines r, s is determined by solving a 4×3 linear system of equations,
that is a system of 4 equations in 3 unknowns. If the direction vectors of the lines → −
v1 and
→
−
v2 are parallel, so are the lines, and no intersection exists, thus no solution exists of the
system; note that r and s lie in the same plane, that is they are coplanar However, when the
direction vectors are not proportional, a new phenomenon occurs. If the lines are coplanar,
then they must intersect, exactly one common point exists, and the system has exactly one
solution. But, if the lines are not coplanar, then no common points exist, and the system has
no solutions; r and s are called skew lines.
Skew lines are a new important feature of three space compared with two space. In the plane,
two lines are either intersecting or they are parallel, notice that they of course lie in the same
plane! In three space, though, two lines can also be not coplanar: in this case they are not
intersecting and they are not parallel and we call them skew lines.
Exercise. How do you find the plane containing two parallel lines?

Exercise. Can you find a plane containing two skew lines?

Actually, when intersecting a line and a plane, this can be done more efficiently without us-
ing the 3×3 system equation. We can just use the parametric equation of the line, containing
only one variable t, and substitute into the equation of the plane.
Example. Find the intersection of the plane α : x + y + z = 0 and of the line

y−x=1
r:
z−y =1

We first find a parametric equation of the line. We can do this by finding the direction vector
using the vector product and then picking a random point of the line. However, in this case it
is simpler to use the structure of the equations. If we set y = t we can readily solve the system
and get the parametric equation
     
x −1 1
 y  =  0  + t 1 
z 1 1

To find r ∩ α we substitute the parametric equation of the line in the equation of the plane and
we get
(t − 1) + (t) + (t + 1) = 0
and thus t = 0. Setting t = 0 in the parametric equations of r we get the intersection point
(−1, 0, 1).

Exercise. Find the parametric equations for r using the vector product. Do you get the same
equations?

L4.3 Orthogonal projection and symmetries. If a point P does not lie on a plane α, then
the orthogonal proiection of the point on the plane is the point of α closest to P , that is it
the best approximation of P in α. This is why orthogonal projections naturally appears in
numerical applications. Symmetries (or reflections), besides their aesthetical appeal, have
similar connections with applications.
Let’s see how to find orthogonal projections on planes.
Lemma. Fix O ∈ α. The orthogonal projection of P on α is the point H such that

−−→ −−→ − −
n→
HP = OP · n→
α
α − .
||n→α ||
2

Proof. Here is the basic remark: for a given point P , let H ∈ α be the orthogonal projection
of P on α. Then the triangle of vertices P, O and H has a right angle in H and we have the
−−→ −−→ −−→
following vector equation OH + HP = OP .
−−→ −−→
Thus, HP is the orthogonal projection of OP on − n→ α and this is enough to conclude the
proof. QED

Example. Consider the plane α : z = 0 and the point P (x, y, z) . To find the orthogonal projec-
→
−
tion of the point P , we pick O(0, 0, 0) and then we apply the lemma noticing that −
n→
α = k and
we get
−−→ →
− →
− →
− → −→ − →
−
HP = (x i + y j + z k ) · k k = z k
and we get H(x, y, 0) as it should be.

Exercise. If P (xP , yP , zP ) and α : ax + by + cz = d , let

axP + byP + czP − d

(xP − aT, yP − bT, zP − cT ).

Prove this formula.

Rather than using the previous formulae, it is possible to follow an alternative geometric
argument in order to find orthogonal projections. Given a point P and a plane α, let H ∈ α
be the orthogonal projections of P . In order to find H , find the unique line r passing
through P and such that r ⊥ α. Then, H = r ∩ α.
Example. Find the orthogonal projection of P (3, 2, 1) on the plane a : x + y + z = 0. The
direction vector of the line r can be chosen equal to the normal vector of the plane, that is
→
− −→ →
− → − → −
p =n α = i + j + k . Since P ∈ r the parametric equations of the line are

     
x 3 1
 y  =  2  + t 1 
z 1 1

To find r ∩ α we substitute the parametric equation of the line in the equation of the plane
and we get
(3 + t) + (2 + t) + (1 + t) = 0
and thus t = −2. Setting t = −2 in the parametric equations of r we get the orthogonal
projection H(1, 0, −1).
The treatment of symmetries is very similar to the one of orthogonal projections. The key
remark is the following: if Q is the symmetric of P with respect to the plane α, then the
−−→
middle point of P Q is H where H is the orthogonal projection of P on α.
Lemma. Fix O ∈ α. The symmetric of P with respect to the plane α is the point Q such
that
−−→ −−→ → − n→
QP = 2OP · −
α
nα − → .
||nα ||2

Example. Find the point Q the symmetric point of P (1, 2, 3) with respect to the plane α : z = 0.
Fix O(0, 0, 0) and note that −
n→
α = (0, 0, 1). Let Q(x, y, z) and apply the Lemma:

−−→ −
−−→ OP · n→ α−
QP = 2 − n→
α = 2(1, 2, 3) · (0, 0, 1)(0, 0, 1) = (0, 0, 6),
||n→
α ||
2

−−→
thus QP = (P − Q) = (1 − x, 2 − y, 3 − z) = (0, 0, 6) and hence Q(1, 2, −3).

Exercise. Find an expression for the coordinates of Q as we did for the coordinates of the
orthogonal projection.

L4.4 Further exercises.

1. Find the orthogonal projection of the points A(1, 2, 3) and B(3, 2, 1) on the plane α :
x − z = 1.

2. Describe the mutual position of all possible pairs chosen from:

x+y+z =0
α : x + y + z = 1, r : x = y = z, s : { .
x−y−2=0

3. Choose three pairwise non-parallel planes and consider the three lines obtained as the
pairwise intersections of the planes. Do the line have a point in common? Does the answer
(YES/NO) depend on the initial choice of the planes?

4. Consider the triangle T of vertices A(1, 0, 0), B(0, 1, 0), C(0, 0, 1), and let T 0 be the or-
thogonal projection of T on the plane x + y + z = 5. Find the area of T 0 . How does it
compare to the area of T ?

5. Same exercise with the plane π : x + y − z = 5.

6. Discuss symmetries with respect to a line.

7. Find three planes intersecting exactly in the point P (1, 2, 3). Can you find three lines
intersecting exactly at the same point?