I'm stuck with problem supposed to be trivial.
I need to find tangent line witch touches curve $y^2 = -4ax$ at the point $(x_0,y_0)$

Rewriting it as $$x = -\frac {y^2}{4a}$$ Taking derivative: $$f(y)'=-\frac {y}{2a}$$ Then slope is $$f(y_0)'=-\frac {y_0}{2a}$$ Then make a substitution in tangent line equation at the point $(x_0,y_0)$ to find the offset:
$$x = ky+b \iff b = x-ky \iff b = x_0 -\frac {y_0^2}{2a}$$ Then write our tangent line equation: $$x = -\frac {y_0y}{2a} + x_0 -\frac {y_0^2}{2a} \iff 2ax = -y_0y + 2ax_0 -y_0^2$$

But this result doesn't match with result given in the book: $$4ax+2yy_0-y_0^2 = 0$$ Help me out, guys.


You computed $\frac{dx}{dy}$ at $(x_0,y_0)$. This is the reciprocal of the slope. Easy fix, the slope at $(x_0,y_0)$ is $-\frac{2a}{y_0}$.

Alternately, differentiate implicitly. We get $$2y\frac{dy}{dx}=-4a.$$ So the slope of the tangent line at $(x_0,y_0)$ is $-\frac{4a}{2y_0}$.

Remark: Your procedure should also have worked. However, there is a mistake in the next to last displayed line. It should be $2ax_0$, not $2x_0$. (When you cleared denominators by multiplying through by $2a$, the term $x_0$ did not get multiplied by $a$.)

Edit: The error has been corrected. To get the answer in the book, note that $2ax_0=-\frac{y_0^2}{2}$.

  • $\begingroup$ Thank you for the answer. But, for starters, i don't see why I have to flip my derivative. As it obvious from my question, I look at equation like y is a dependant variable (Yes, it's non-conventional, but can't take square root here and solve for x). So, why would I flip then? And even if I flip, I get $x = - \frac {2ay} {y_0} + x_0 -2a$ $\endgroup$
    – Dmitri K
    May 10 '14 at 16:47
  • $\begingroup$ Agree with your remark. I lost an a. Already made the edit, thanks. $\endgroup$
    – Dmitri K
    May 10 '14 at 16:49
  • $\begingroup$ You are welcome. $\endgroup$ May 10 '14 at 16:50
  • $\begingroup$ What about my previous comment? Your suggestion gives result that also doesn't match to result given in the book. Honestly I don't see why it would be valid to do it your way. Please, review my previous comment. $\endgroup$
    – Dmitri K
    May 10 '14 at 16:52
  • $\begingroup$ Your answer is now correct. To get the book's form, as my edit at the end suggests, use the fact that $2ax_0=y_0^2/2$. $\endgroup$ May 10 '14 at 16:54

There are a couple of mistakes in your working out (I see the second one has now been corrected in your question)

1. Where you write

$$b = x-ky \iff b = x_0 -\frac {y_0^2}{2a}$$

you should have instead

$$b = x-ky \iff b = x_0 +\frac {y_0^2}{2a} $$

2. In addition you forgot a factor of $a$ on the $x_0$ so

$$x = -\frac {y_0y}{2a} + x_0 +\frac {y_0^2}{2a} \iff 2ax = -y_0y + 2x_0 +y_0^2$$


$$x = -\frac {y_0y}{2a} + x_0 +\frac {y_0^2}{2a} \iff 2ax = -y_0y + 2ax_0 +y_0^2$$

Finally, you can just substitute for $2ax_0$ using $y_0^2=-4ax_0$, giving

$$ 2ax = -y_0y - \frac{y_0^2}{2} +y_0^2 $$

from which the result follows

$$ 4ax + 2y_0y - y_0^2 = 0 $$


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.