I am studying the proof of the following proposition in Hartshorne - Let $k$ be an algebraically closed field. There is a natural fully faithful functor $Var(k)\longrightarrow Sch(k)$ from the category of varieties over $k$ to schemes over $k$.

First they make some statements about general topological spaces (in which I have a doubt).

Let $X$ be any topological space. And let $t(X)$ be the set of non-empty irreducible closed subsets of $X$. If $Y, Y_1, Y_2, Y_i$ $(i\in I)$ are closed subsets of $X$, then we can check that $t(Y)\subset t(X)$, $t(Y_1)\cup t(Y_2)=t(Y_1\cup Y_2)$ and $\bigcap_{i\in I}t(Y_i)=t(\bigcap_{i\in I}Y_i\big)$. From this, we can make $t(X)$ a topological space, by taking closed subsets of $t(X)$ to be subsets of the form $t(Y)$, where $Y$ is a closed subset of $X$. Further if $f:X_1\longrightarrow X_2$ is a continuous, we obtain a map $t(f):t(X_1)\longrightarrow t(X_2)$ by sending $Y\in t(X_1)$ to $cl(f(Y))$ in $X_2$. Thus $t$ is a functor on topological spaces.

Further, we define a map $\alpha:X \longrightarrow t(X)$, by $\alpha(P)=cl\{P\}$. We see that $\alpha$ is continuous because for any closed subset $t(Y)$ of $t(X)$, $\alpha^{-1}(t(Y))=Y$, which is closed in $X$. Next they make the following statement :

$\alpha$ induces a bijection between the set of open subsets of $X$ and the set of open subsets of $t(X)$. This is where I have doubts.

I think it is enough to prove that there is a bijection between the corresponding closed sets. What I have so far is : if $Y$ is a closed subset of $X$, then $\alpha^{-1}(t(Y))=Y$, and $\alpha(Y)\subset t(Y)$. Also if $Y\neq Z$ closed subsets of $X$, I can show that $\alpha(Y)\neq\alpha(Z)$. What more is needed to prove that there is a bijection?


Let $U=X\setminus Y$ be an open subset of $X$ (of course $Y$ is closed). The induced map can be defined as $$ \alpha(U) = t(X)\setminus t(Y), $$ which is well defined since $\alpha(U)$ is an open subset of $t(X)$ by definition of its topology. This map is obviously a bijection between the open subsets of $X$ and the open subsets of $t(X)$.

  • $\begingroup$ We already have a map $\alpha:X\longrightarrow t (X) $. Is the $\alpha $ you have defined, induced by it? $\endgroup$ – gradstudent Dec 12 '13 at 18:30
  • $\begingroup$ Yes, this is what I mean. The original $\alpha$ is defined on points, the induced one on open subsets, but they are morally the same map $\endgroup$ – Abramo Dec 12 '13 at 18:36
  • $\begingroup$ Thank you! I understand it now. $\endgroup$ – gradstudent Dec 13 '13 at 2:42

I'm not sure Abramo's answer above is correct as described here.

To be clear, we can define a bijection from closed sets of $X$ to closed sets of $tX$ by sending a closed set $Y \subseteq X$ to the closed set $tY \subseteq tX$. This map is clearly subjective by definition of the topology on $tX$. As was mentioned above, we have $\alpha^{-1}(tY) = Y$, for any closed $Y \subset X$, and this gives injectivity.

Closed set and open sets are in bijection by complementation so the above bijection can be made to be between open sets by sending an open $X-Y$ in $X$ to $tX - tY$ in $tX$.


Let me elaborate more on Abramo's answer. In fact we are looking at $\alpha^{-1}$ for the correspondence of open subsets.

Let $Y\subset X$ be closed so that $X-Y$ is open. Consider $\alpha^{-1}(tX-tY) = X-\alpha^{-1}(tY)$.

What is $\alpha^{-1}(tY)$? It is just $Y$: for $P\in X$, $\alpha(P) =\overline{\{P\}}$ is an irreducible subset of $Y$ if and only $P\in Y$! (Note that the closure of a point is always irreducible).

Thus, we get a bijection between open sets of $X$ and open sets of $tX$, by

$$tX-tY \longleftrightarrow X-Y.$$


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.