Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

So, this seems like it should be fairly simple, right? $$ F(x,y) = {xy\over \sqrt{x^2+y^2}} $$

If one takes $\lim_{(x,y) \to (0,0)}F(x,y)$ along the line $y=x$, the limit simplifies to $ \lim_ {x \to 0}\frac{x^2}{\sqrt{x^2+x^2}} $ which quickly reduces to $|x|/\sqrt{2}$, ie 0.

If one takes $\lim_{(x,y) \to (0,0)}F(x,y)$ along the parabola $y=x^2$, the equation simplifies to $\lim_{x \to 0} \frac{x^3}{\sqrt{x^2+x^4}}$ which evaluates to $ 0 $ not as I thought $ \infty $ by multiplication by $\frac{1 \over \sqrt{x^6}}{1 \over \sqrt{x^6}}$.

Therefore the limit could (as opposed to not, as I originally thought) exist. Alas, according to the book, this is incorrect - the limit is 0 (of course).

What is invalid about my approach and how should the problem be approached?

Thanks in advance!

EDIT: I really appreciate all the answers, it has given me much to look into given time.... Dan's was most direct to the skillset this section is trying to develop.

share|cite|improve this question
Your evaluation along the parabola is wrong. – DonAntonio Mar 14 '13 at 5:07
Sweet..... but I keep going over it, and I can't find anything wrong with my evaluation of the parabolic limit by multiplication .... anyone so kind as to enlighten me? :\ – user1833028 Mar 14 '13 at 5:28
Hint: $\sqrt{x^2+x^4} \sim \sqrt{x^2}$ – Jean-Claude Arbaut Mar 14 '13 at 6:19
Note that $\sqrt{2x^{2}}=|x|\sqrt{2}$ and not $x\sqrt{2}$. – T. Eskin Mar 14 '13 at 7:12
AAAAAAAAAhhhhhhhhh...... $1/\infty = 0$. Whoops. – user1833028 Mar 14 '13 at 7:26
up vote 5 down vote accepted

$$\frac{x^3}{\sqrt{x^2+x^4}}=\pm\frac{x^2}{\sqrt{1+x^2}}\xrightarrow[x\to 0]{}0$$

In fact, check that


share|cite|improve this answer
You're on fire tonight, @Donantonio – Rustyn Mar 14 '13 at 5:07
I'm on insomnia tonight, Rustyn...*sigh*...But in a little while more I'm going to hit the sack! – DonAntonio Mar 14 '13 at 5:12
I know how that goes. – Rustyn Mar 14 '13 at 5:15

You can do it all at once using $|xy|\leq 2|xy|\leq x^2+y^2$ which is an often useful estimate following from $x^2+y^2-2|x||y|=(|x|-|y|)^2\geq 0$. Then $$ 0\leq |F(x,y)|\leq \frac{x^2+y^2}{\sqrt{x^2+y^2}}=\sqrt{x^2+y^2}. $$ So $$ \lim_{(x,y)\rightarrow(0,0)} F(x,y)=0 $$ follows by squeezing.

share|cite|improve this answer

To find the limit, write $x=r\cos t$, $y=r\sin t$.

share|cite|improve this answer
if we take two path,like x goes to zero and y goes to zero,we see that limit exist,and both goes to zero,except where both x and y goes to zero,i think limit in this case exist – dato datuashvili Mar 14 '13 at 5:10
Hmm.... book does not mention any method like this... do you know of any link that I could use to see an example of this, sir? – user1833028 Mar 14 '13 at 5:12
If you make the substitution I suggested, the expression simplifies to $r\sin t\cos t$. This approaches $0$ as $(x,y)\to 0$, since $r\to 0$. – André Nicolas Mar 14 '13 at 5:12
listen to @AndréNicolas Nicolas – dato datuashvili Mar 14 '13 at 5:13
@user1833028: Sorry, I do not know an explicit reference. It is quite standard for the denominator $x^2+y^2$ and relatives. – André Nicolas Mar 14 '13 at 5:14
  1. I would be more careful about your logic for the "parabola": The manipulations you've suggested are suited for a limit at infinity, not at zero.

  2. The proof can be tricky, since you have to deal with two variables at once. My suggestion: Let $m = \max\{x,y\}$. Certainly $m \rightarrow 0$ as $x,y \rightarrow 0$. Can you find an expression, using $m$, which is greater than or equal to $f(x,y)$, and then show that this expression goes to zero?

share|cite|improve this answer

We should prove that: $$\forall\epsilon>0, \exists\delta>0,\forall (x,y): \left(0<||(x,y)-(0,0)||<\delta\longrightarrow\big|\frac{xy}{\sqrt{x^2+y^2}}-0|<\epsilon\right)$$ Firstly, see that $||(x,y)-(0,0)||<\delta$ is equvalant to saying that $\sqrt{x^2+y^2}<\delta$ and so both of $|x|$ and $|y|$ should be lass than $\delta$, so if we set $$z=\text{max}(|x|,|y|)$$ we get $$\big|\frac{xy}{\sqrt{x^2+y^2}}-0|=\frac{|x||y|}{\sqrt{x^2+y^2}}\leq\frac{zz}{\sqrt{z^2+0}}=z<\epsilon$$ Hence we need enough to take $\delta=\epsilon$.

share|cite|improve this answer
Very nicely done! +1 – amWhy Mar 15 '13 at 1:31

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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