Let $X = V(xy-zw) \subset \mathbb{P}^3$ (the variables are $x,y,z,w$). I know from various sources that $Pic(X) \cong \mathbb{Z} \oplus \mathbb{Z}$, where generators are the two lines $l_x = V(x,w) \subset X$ and $l_y = V(y,w)$. (see for example, Hartshorne, II, Example 6.6.1,2).

I can show that $Pic(X)$ is generated by $\{l_x,l_y\}$, but I'm having trouble at showing that the kernel of $\mathbb Z l_x \oplus \mathbb Z l_y \to Pic(X)$ is trivial. It seems to me that to show this it is equivalent to show that for any $f \in K(X)$, the associated divisor $(f)$ has coefficient zero at prime divisors $l_x$ and $l_y$. In turn, this amounts to showing that $v_{l_x}(f) = 0 = v_{l_y}(f)$.

However, I cannot understand why this would be true. More precisely, it seems that if I put $f = x/y$, then $f$ is a non-zero rational function on $X$, and $v_{l_x}(f) \geq 1$ since $f$ is zero on the whole $D(y)\cap l_x$, whence in the maximal ideal of $\mathcal O _{l_x}$.

Thanks for your help. Also I don't know to what extent this is a trivial/stupid question, so please be indulgent if that would be the case..


1 Answer 1


No, you don't need to show that, for every $f\in K(X)$, $\operatorname{div}(f)$ has $0$ coefficient at $l_x$ and $l_y$. What you need to show is that, if $\operatorname{div}(f)\in\mathbb{Z}l_x\oplus\mathbb{Z}l_y$, then $\operatorname{div}(f)=0$. It is slightly different.

In fact, $\operatorname{div}(x/y)=V(x,w)+V(x,z)\notin\mathbb{Z}l_x\oplus\mathbb{Z}l_y$.

  • $\begingroup$ Indeed, you are right. Thank you so much! $\endgroup$
    – Olivier
    May 12, 2015 at 7:13

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.