MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

More precisely, if $\mathcal F_i$ is a system of sheaves, is it the case that $$ (\lim \mathcal F_i)_p = \lim ((\mathcal F_i)_p) $$ and similarly for colimits? I can see how to get a map $$ (\lim \mathcal F_i)_p \rightarrow \lim ((\mathcal F_i)_p) $$ by taking the stalks in the diagram for $\lim \mathcal F_i$ and then using the universal property of $(\lim \mathcal F_i)_p$, but I can't see how I should use the sheaf condition to go the other way.

I learning algebraic geometry and taking a class with Hartshorne's book, and we showed that a map of sheaves is injective/surjective iff it is injective/surjective on stalks. Since one can check injectivity/surjectivity by computing the kernal/cokernal, I am thinking that what I wrote above would be a generalization of this fact.

Looking at sheafication as a adjoint and how it interacts with limits and colimits really made me feel more comfortable with sheafication, so I'm hoping that looking the interaction between sheaves and their stalks in terms of limits and colimits will give me some intuition.

share|cite|improve this question
Yes, Hartshorne's book is not too good for this type of fundamental categorical fact. The answer given by Martin is very familiar to category theorists; see for instance the book by Mac Lane and Moerdijk on topos theory (where they discuss sheaves over a space). You can also think of the process of taking a stalk as a particular kind of colimit, which always commutes with colimits. It commutes with finite limits because taking a stalk is an example of a filtered colimit (see Categories for the Working Mathematician). – Todd Trimble Nov 7 '11 at 19:24
I recommend George Kempf's little paperback, Algebraic Varieties. The (true) result you want is Lemmna 4.4.1. Kempf covers in 140 pages sheaves, varieties, cohomology, and RRT for curves and surfaces. – roy smith Nov 8 '11 at 5:13
up vote 6 down vote accepted

Let $F$ be a sheaf on $X$ and $p \in X$. Then $F_p$ is just the pullback $i^{-1} F$, where $i : \{p\} \to X$ is the inclusion of a point. Now $i^{-1}$ is left adjoint to $i_*$, thus cocontinuous, i.e. preserves all colimits. This shows that the canonical morphism $\mathrm{colim}_i(F_p) \to \mathrm{colim}_i(F)_p$ is an isomorphism.

Now for limits, we get a canonical morphism $(\mathrm{lim}_i F)_p \to \mathrm{lim}_i(F_p)$. This is almost never an isomorphism (neither injective nor surjective). Consider infinite products and see what happens. However, the (left) exactness of $F \mapsto F_p$ means that this functor preserves finite limits.

share|cite|improve this answer
Thanks, I think this is exactly what I was looking for. – Drew Nov 7 '11 at 19:30

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.