When telling us the algebra of limit for functions we were told that,

Let $f,g :D(\subseteq \mathbb{R}) \to \mathbb{R}$ such that $a$ is a limit point of $D$ and also such that $g(x)\ne 0$ for all $x\in D$ and $\displaystyle\lim_{x\to a}g(x)\ne 0$ then, $$\displaystyle\lim_{x\to a}\dfrac{f(x)}{g(x)}=\dfrac{\displaystyle\lim_{x\to a}f(x)}{\displaystyle\lim_{x\to a}g(x)}$$ Assume that the values all limits are finite provided they exist.

Is "for all" (see the bold text) here necessary? For example, if $a\in D$ and $g(a)$ is defined then I don't think that it is necessary that $g(a)\ne 0$. In this case I think that it will be sufficient to say that, $$(g(x)\ne 0\ \forall x\in D\setminus \{a\})\land(\displaystyle\lim_{x\to a}g(x)\ne 0) $$ Am I wrong somewhere?


It is sufficient that $g(x)$ be nonzero in some open set at $a$. Equivalently in an open interval at $a$. If $g$ is continuous this is equivalent to saying $g(a) \ne 0$. However we must have $\displaystyle\lim _{x \to a} g(x) \ne 0$ or the RHS may not exist. If $g$ is continuous this is equivalent to demanding $g(a) \ne 0$.

Oh, so you say, the equality still holds if the LHS fails to exist as well. But suppose $f(x) = g(x) = x$ for all $x \in D = \mathbb R$ and let $a = 0$. Then $f(x)/g(x) = x/x = 1$ at every point. So the LHS is $\lim _{x \to 0} 1 = 1$ which certainly exists. But the RHS evaluates to $0/0$ which is undefined.

Oh, but you further say, what if we choose to evaluate $0/0$ as $1$ for the purposes of this formula? Suppose now that $f(x) =2x$ and $g(x) = x$ for all $x \in D = \mathbb R$ and let $a = 0$. Then $f(x)/g(x) = 2x/x = 2$ at every point. So the LHS is $\lim _{x \to 0} 2 = 2$. But the RHS again evaluates to $0/0$ which we may evaluate as $1$ as we decided. But then the formula becomes $2=1$ which is not true.

It is possible that if $g$ is not continuous we can have $g(a) = 0$ and the formula still holds. For example let $g(x) = 1$ for all $x \ne 0$ and $g(0)=1$. Again let $f(x) = x$ and $a = 0$. The the function $f(x)/g(x) = x$ when $x \ne 0$ and is undefined at $x=0$. However we can still talk about its limit at $0$, which turns out to be $0$. Moreover $\lim f(x) = 0$ and $\lim g(x) = 1$. So the LHS exaluates to $0/1=0$ which is equal the $RHS$.

  • $\begingroup$ In the second paragraph, the function $g(x)=x$ doesn't satisfy $\displaystyle\lim_{x\to 0}g(x)=0$ and similar is the case for the third paragraph. Furthermore, when you say "...we must have $g(a)\ne 0$ or the RHS mayn't exist.", do you mean when $g$ is continuous or always? $\endgroup$ – user 170039 Apr 10 '15 at 13:09
  • $\begingroup$ To address the first point: In the second and third paragraphs we do have $\displaystyle \lim_{x \to 0} g(x) = 0$. We defined $g(x) = x$ so by definition $g(x) \to 0$ as $x \to 0$. $\endgroup$ – Daron Apr 10 '15 at 13:52
  • $\begingroup$ Sorry, it was a typo. I meant $\displaystyle\lim_{x\to 0}g(x)\ne 0$. $\endgroup$ – user 170039 Apr 10 '15 at 14:37
  • $\begingroup$ I have chosen $\lim g(x) = 0$ so as to illustrate that $\lim g(x) = 0$ is a necessary assumption to make the formula work. $\endgroup$ – Daron Apr 10 '15 at 15:57

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.