This question concerns a reduction argument that occurs in the proof of Theorem I.5.3 in Hartshorne. In particular, let $Y$ be an affine variety of dimension $r$ in $\mathbb{A}^n$. Then by (4.9), $Y$ is birational to a hypersurface $X$ of $\mathbb{P}^{r+1}$ (Hartshorne writes $\mathbb{P}^{n}$, I believe this is a typo). Then Hartshorne writes

"Since birational varieties have isomorphic open subsets, we reduce to the case of a hypersurface. It is enough to consider any open affine subset of $Y$, so we may assume that $Y$ is a hypersurface in $\mathbb{A}^n$, defined by a single irreducible polynomial $f(x_1,\dots,x_n) = 0$."

Question 1: I wonder if by the statement "birational varieties have isomorphic open subsets," Hartshorne really means "birational varieties $X, Y$ have open sets $U_X, U_Y$ that are isomorphic" (which is Corollary I.4.5(ii)). The way he writes it, it could mean that for any open set $U_X$ there exists an open set $U_Y$ such that $U_X, U_Y$ are isomorphic. Is this latter statement true? Can somebody please clarify this?

Question 2: Why can we assume that $Y$ is a hypersurface in $\mathbb{A}^{r+1}$ (I believe this is another typo since Hartshorne writes $\mathbb{A}^n$)? Hartshorne's argument seems unclear.


Q1: As noted in the comments, your interpretation is correct. It's not true that given $X$ birational to $Y$, for every $U_X \subset X$ open there exists an open $U_Y \subset Y$ so that $U_X \simeq U_Y$. Indeed, consider a variety with a point singularity (for instance a nodal cubic) and its desingularization. These varieties are certainly birational, but no open set containing the singularity can be isomorphic to an open subset of the nonsingular variety.

Q2: In the proof of I.5.3, which asserts that the set of singular points of a variety of $Y$ is a proper closed subset, it is first shown that the set of singular points is a closed subset (using affine covers). Reduction to an affine hypersurface is used to show that the set of singular points is a proper subset of $Y$. Note that $Y$ is birational to a hypersurface in $\mathbb P^{r+1}$, and an open subset of this projective hypersurface is isomorphic to a hypersurface in $\mathbb A^{r+1}$. This implies that an open subset $U$ of $Y$ is isomorphic to an open subset of $V$ of an affine hypersurface.

From these observations, it suffices to show that the subset of singular points of the affine hypersurface is a proper subset. Given this, its complement is open by the first part of the proof, and therefore intersects $V$ nontrivially. Since $U \subset Y$ is isomorphic to $V$, it follows that $Y$ contains non-singular points and the set of singular points is a proper subset.

  • 1
    $\begingroup$ @Manos, just intersect the hypersurface with an affine open set which intersects it non-trivially. $\endgroup$ – Mariano Suárez-Álvarez Jul 25 '14 at 16:09
  • $\begingroup$ You can choose an appropriate affine open chart for $\mathbb P^n$, and intersect it with your hypersurface. $\endgroup$ – vociferous_rutabaga Jul 25 '14 at 16:10
  • $\begingroup$ Well, try to prove it :-) $\endgroup$ – Mariano Suárez-Álvarez Jul 25 '14 at 16:24
  • $\begingroup$ @Manos you simply need that an open subset of $Y$ is isomorphic to an open subset of an affine hypersurface -- see the clarifications that I've made above. This only requires that some open subset of a projective hypersurface be an affine hypersurface. I've tried to clarify this in my answer. $\endgroup$ – vociferous_rutabaga Jul 25 '14 at 16:33

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.