MHB Can three tangents to a circle meet at a common point?

  • Thread starter Thread starter poissonspot
  • Start date Start date
  • Tags Tags
    Circle
AI Thread Summary
The discussion centers on the possibility of three tangents to a circle meeting at a common point, with the conclusion that such a scenario leads to contradictions. The analysis involves using a unit circle and a point outside it to derive the equations for tangent lines. By examining the discriminant of the resulting quadratic equation, it is shown that there are always two tangent lines for a point outside the circle. The conversation also touches on alternative methods, such as vector algebra, to explore the problem further. The participants express interest in additional solutions to the question.
poissonspot
Messages
39
Reaction score
0
I considered the question of whether three tangents to the circle could meet at a common point and only came up with a contradiction by lengthy constructive means.
Circles are "nice", so there must be some clever ways of showing this fact that given a point outside the circle there are two tangent lines to the circle passing through this point. Any ideas? Thanks mucho,
 
Mathematics news on Phys.org
You could, without loss of generality, use a unit circle, and a point on the $y$-axis outside of the circle $(0,b)$ where $1<b$.

The equation of the circle is:

$$x^2+y^2=1$$

The equation of the tangent lines is:

$$y=mx+b$$

Substitute for $y$ from the line into the circle:

$$x^2+(mx+b)^2=1$$

$$x^2+m^2x^2+2bmx+b^2=1$$

$$\left(m^2+1 \right)x^2+(2bm)x+\left(b^2-1 \right)=0$$

Now, by analyzing the discriminant, we find:

$$\Delta=(2bm)^2-4\left(m^2+1 \right)\left(b^2-1 \right)$$

$$\Delta=4\left(b^2m^2-b^2m^2+m^2-b^2+1 \right)$$

$$\Delta=4\left(m^2-b^2+1 \right)$$

We see that with $1<b$, there will be two real values of $m$, given by:

$$m=\pm\sqrt{b^2-1}$$

This implies there are two tangent lines:

$$y=\pm\sqrt{b^2-1}x+b$$
 
MarkFL said:
You could, without loss of generality, use a unit circle, and a point on the $y$-axis outside of the circle $(0,b)$ where $1<b$.

The equation of the circle is:

$$x^2+y^2=1$$

The equation of the tangent lines is:

$$y=mx+b$$

Substitute for $y$ from the line into the circle:

$$x^2+(mx+b)^2=1$$

$$x^2+m^2x^2+2bmx+b^2=1$$

$$\left(m^2+1 \right)x^2+(2bm)x+\left(b^2-1 \right)=0$$

Now, by analyzing the discriminant, we find:

$$\Delta=(2bm)^2-4\left(m^2+1 \right)\left(b^2-1 \right)$$

$$\Delta=4\left(b^2m^2-b^2m^2+m^2-b^2+1 \right)$$

$$\Delta=4\left(m^2-b^2+1 \right)$$

We see that with $1<b$, there will be two real values of $m$, given by:

$$m=\pm\sqrt{b^2-1}$$

This implies there are two tangent lines:

$$y=\pm\sqrt{b^2-1}x+b$$

Thanks, this is close to how I approached the problem. I used vector algebra though instead and solved for points where the dot product was equal to zero.

btw, I'm still interested in other solutions if anyone else is. Thx again,
 
Seemingly by some mathematical coincidence, a hexagon of sides 2,2,7,7, 11, and 11 can be inscribed in a circle of radius 7. The other day I saw a math problem on line, which they said came from a Polish Olympiad, where you compute the length x of the 3rd side which is the same as the radius, so that the sides of length 2,x, and 11 are inscribed on the arc of a semi-circle. The law of cosines applied twice gives the answer for x of exactly 7, but the arithmetic is so complex that the...
Back
Top