6
$\begingroup$

In Görtz and Wedhorn's Algebraic Geometry I, there's the following proposition:

Proposition 3.4. Let $(X,\mathcal O_X)$ be a locally ringed space. If $Y$ is an affine scheme then the natural map below is an isomorphism $$\mathsf{Hom}(X,Y)\overset{\cong}{\longrightarrow} \mathsf{Hom}(R,\Gamma(X,\mathcal O_X))$$

A proof is given only for the case where $X$ is a scheme, and proceeds by reducing to the anti-equivalence of affine schemes with commutative rings and then gluing the morphisms.

For the general case, the authors cite Prop 1.6.3 of the 1971 edition of EGA I, to which I don't have access. Can anyone reproduce/explain the proof?

Improve this question
$\endgroup$
5
  • 2
    $\begingroup$ The 1971 edition is in Springer Grundlehren series, which is widely available. $\endgroup$
    – abx
    Jan 11, 2016 at 13:30
  • $\begingroup$ @abx widely available as it may be, I don't have access to it and I haven't been able to find it online. Also, will the downvoter explain the downvote? $\endgroup$
    – Arrow
    Jan 11, 2016 at 13:34
  • 1
    $\begingroup$ Injectivity of the map follows from the characterization of the diagonal in $Y\times Y$ as the zero scheme of the kernel of the multiplication morphism $R\otimes R \to R$. For surjectivity, first you construct the set map $f:X\to Y$ by sending $p\in X$ to the kernel of $R\to \Gamma(X,\mathcal{O}_X) \to \kappa(p)$, then you check that $f^{-1}(D(r))$ is open for each $r\in R$, and finally you define $f^\#:\mathcal{O}_Y \to f_*\mathcal{O}_X$ to be the unique morphism of sheaves of algebras compatible with $\widetilde{R} \to \widetilde{\Gamma(X,\mathcal{O}_X)}$ on $Y$. $\endgroup$
    – Jason Starr
    Jan 11, 2016 at 13:35
  • 1
    $\begingroup$ See also the elusive EGA I.1.8 which is found in... EGA II at the end as an addendum, which is freely available online. :) $\endgroup$
    – Dylan Wilson
    Sep 23, 2016 at 13:46
  • $\begingroup$ See stacks.math.columbia.edu/tag/01I1. $\endgroup$
    – Jérôme Poineau
    Sep 23, 2016 at 14:15

1 Answer 1

Reset to default
6
$\begingroup$

Here is a sketch: Let $\alpha : R \to \Gamma(X,\mathcal{O}_X)$ be a ring homomorphism. We want to define a morphism $f:X \to \mathrm{Spec}(R)$ which is $\alpha$ on global sections. Let $x \in X$, and consider the composition of $\alpha$ with $\Gamma(X,\mathcal{O}_X) \to \mathcal{O}_{X,x}$, $s \mapsto s_x$. Pull back the unique maximal ideal $\mathfrak{m}_x$ of $\mathcal{O}_{X,x}$ to $R$, this is a prime ideal of $R$. Denote this by $f(x)$. This defines a map of sets $f : X \to \mathrm{Spec}(R)$. To show continuity, let $r \in R$ and observe that the preimage of $D(r) \subseteq \mathrm{Spec}(R)$ in $X$ is the set of points $x \in X$ such that $\alpha(r)_x$ is invertible in $\mathcal{O}_{X,x}$. This set $D(\alpha(r))$ is easily seen to be open. Also notice that by the universal property of localization, $\alpha$ induces a homomorphism $R[r^{-1}] \to \Gamma(D(\alpha(r)),\mathcal{O}_X)$, i.e. $\Gamma(D(r),\mathcal{O}_{\mathrm{Spec}(R)}) \to \Gamma(f^{-1}(D(r)),\mathcal{O}_X)$. These maps are compatible and therefore glue to a morphism of sheaves $f^{\#}:\mathcal{O}_{\mathrm{Spec}(R)} \to f_* \mathcal{O}_X$.

Improve this answer
$\endgroup$

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.