Relation between direction cosines - l² + m² + n² = 1: Question bank

The unit-sphere picture (anchor for everything below)

Figure — Relation between direction cosines - l² + m² + n² = 1

The projections onto each axis (the shadows) are exactly $l$ , $m$ , $n$ :

Spot the error

Each prompt contains a plausible wrong step. Name the flaw.

"Angles

6 0^{\circ}, 6 0^{\circ}, 6 0^{\circ}

: cosines

\frac{1}{2}, \frac{1}{2}, \frac{1}{2}

, and

3 \cdot \frac{1}{4} = \frac{3}{4}

, close enough to

1

The error is "close enough."

\frac{3}{4} \neq = 1

exactly, so the triple is impossible — no line makes

6 0^{\circ}

with all three axes. The identity is a strict filter, not an approximation.

"Ratios

(1, 2, 2)

, so direction cosines are

(1, 2, 2)

Missing normalisation. Length is

1 + 4 + 4 = 3

, so the cosines are

(\frac{1}{3}, \frac{2}{3}, \frac{2}{3})

. Raw ratios are never direction cosines unless the length already equals

1

cos^{2} γ = \frac{1}{4}

gives

cos γ = \frac{1}{2}

, so

γ = 6 0^{\circ}

is the only answer."

Dropping the

\pm

. A square root yields

cos γ = \pm \frac{1}{2}

, giving

γ = 6 0^{\circ}

12 0^{\circ}

— the two opposite orientations of the line (both angles lie in the allowed

[0^{\circ}, 18 0^{\circ}]

"Direction cosines of the line through

(1, 1, 1)

and

(2, 2, 2)

: subtract to get

(1, 1, 1)

, divide by

3

, done."

Length is

3

, not

3

. Correct cosines are

(\frac{1}{3}, \frac{1}{3}, \frac{1}{3})

. Dividing by

a^{2} + b^{2} + c^{2}

instead of

a^{2} + b^{2} + c^{2}

is the classic slip.

"A line makes

3 0^{\circ}

with

x

and

4 5^{\circ}

with

y

; then

n = 1 - l^{2} - m^{2}

Confusing

n

with

n^{2}

. The identity gives

n^{2} = 1 - l^{2} - m^{2}

; you must square-root (with

\pm

) to get

n

itself.

"The three direction angles satisfy

α + β + γ = 18 0^{\circ}

, like a triangle."

No such rule. It is the cosines squared that sum to

1

, not the angles that sum to anything fixed. E.g.

54. 7^{\circ}

three times sums to

16 4^{\circ}

, not

18 0^{\circ}

Why questions

Why does the sum of squares equal exactly

1

, rather than the sum of the cosines themselves?

Because

l, m, n

are the perpendicular components of a length-

1

vector, and 3D Pythagoras (Distance Formula in 3D) adds squared components to give squared length. Un-squared sums have no geometric meaning here.

Why do we use cosines to record a direction instead of the raw angles

α, β, γ

Cosines are projections (from the Dot Product

\overset{u}{^} \cdot \hat{i} = cos α

), so they combine linearly and let the angle between two lines become

l_{1} l_{2} + m_{1} m_{2} + n_{1} n_{2}

— see Angle Between Two Lines. Raw angles do not add so cleanly.

Why must we divide direction ratios by their magnitude?

To convert an arbitrary-length direction vector into a unit vector; only then do its components carry the "fraction of one step" meaning that guarantees

l^{2} + m^{2} + n^{2} = 1

(the tip lands on the unit sphere).

Why are there only two independent direction cosines, not three?

The single constraint

l^{2} + m^{2} + n^{2} = 1

removes one freedom: given any two, the third is pinned to

\pm 1 - (\cdot)

. Geometrically, a direction in 3D needs only two angles to specify — a point on a sphere has two coordinates of freedom.

Why does each valid identity solution actually correspond to a real line?

Any

(l, m, n)

with squares summing to

1

is a unit vector, and a unit vector always defines a direction — hence an actual line (and its reverse) through any chosen point via Equation of a Line in 3D.

Why do the two square-root signs give the same line, not two different lines?

(l, m, n)

and

(- l, - m, - n)

point in exactly opposite directions along one and the same straight line. A line has no built-in arrow, so both sign choices describe it.

Edge cases

A line along the positive

z

-axis: what are

l, m, n

α = β = 9 0^{\circ}

γ = 0^{\circ}

, so

(l, m, n) = (0, 0, 1)

. Check:

0^{2} + 0^{2} + 1^{2} = 1

. An axis line has all its "weight" on one component.

Can any two of

l, m, n

both equal

1

No. If

l = m = 1

then

l^{2} + m^{2} = 2 > 1

already, leaving no room for

n^{2} \geq 0

. At most one cosine can reach magnitude

1

(that happens only when the line is an axis).

What happens to the identity as a direction cosine approaches

1

The other two must approach

0

, since

l^{2} \to 1

forces

m^{2} + n^{2} \to 0

. The line is swinging to lie along that axis, its lean onto the other two axes vanishing.

(l, m, n) = (0, 0, 0)

ever valid?

Never. Its squares sum to

0 \neq = 1

; it is the zero vector with no direction. Every legitimate line has at least one non-zero direction cosine.

A line lies along the diagonal of a cube (corner to opposite corner): what are its cosines?

Ratios

(1, 1, 1)

, length

3

, so

(\frac{1}{3}, \frac{1}{3}, \frac{1}{3})

; each angle

\approx 54. 7^{\circ}

. This is the maximally symmetric case where the "one step" splits equally.

If exactly one direction cosine is negative, what does that tell you?

The line leans toward the negative side of that one axis (its angle exceeds

9 0^{\circ}

) while still leaning positively toward the other two. The identity holds regardless of signs — only the squares matter for the sum.

Two lines have cosines

(l, m, n)

and

(- l, - m, - n)

. Are they perpendicular, parallel, or the same?

The same line. Their Dot Product is

- (l^{2} + m^{2} + n^{2}) = - 1

, meaning the angle between the directions is

18 0^{\circ}

— anti-parallel, i.e. one straight line.

Recall One-line summary to keep

$l^{2} + m^{2} + n^{2} = 1$ is Pythagoras on a unit vector: the tip $(l, m, n)$ lives on the unit sphere, angles range over $0^{\circ}$ – $18 0^{\circ}$ , only two cosines are free, and every square root carries a $\pm$ because a line has two opposite directions.

Connections

Direction Ratios — the un-normalised inputs these traps keep confusing with cosines.
Unit Vectors — the source of "squares sum to $1$ ."
Dot Product — why projection = cosine, and why $- 1$ means anti-parallel.
Distance Formula in 3D — the Pythagoras step behind the identity.
Angle Between Two Lines — where $l_{1} l_{2} + m_{1} m_{2} + n_{1} n_{2}$ comes alive.
Equation of a Line in 3D — how a valid $(l, m, n)$ becomes an actual line.

Question bank — Relation between direction cosines - l² + m² + n² = 1

The unit-sphere picture (anchor for everything below)

True or false — justify

Spot the error

Why questions

Edge cases

Connections