Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

The last two pages of Appendix C in Milnor's Characteristic Classes gives an example of a flat bundle with nonzero Euler class. I have a question about the structure group of this bundle.

The example starts with a Riemann surface of genus larger than 2. Its fundamental group, $G$, is represented as a subgroup of $\text{PSL}(2,\Bbb R)$ and acts on the upper half-plane $H$ as the group of covering transformations of the surface.

This action preserves the extended real axis and gives a circle bundle over the surface, $E = (H \times \Bbb{RP}^1)/G \to H/G = S$.

Proof is given that it is isomorphic to the bundle of tangent directions to the surface.

This bundle has a 2-fold cover corresponding to the cohomology class in $H^1(E;\Bbb Z_2)$ that maps into the Thom class of the associated 2-plane bundle under the connecting homomorphism. This class exists because the second Stiefel-Whitney class of the bundle is zero (the surface has even Euler characteristic).

So the 2-fold cover is another circle bundle and each fiber circle in the original is covered by a fiber circle of the two fold cover.

Now it is claimed that the structure group of this two fold cover comes from a representation of $G$ in $\text{SL}(2,\Bbb R)$. I think this means that it is the quotient $(H\times \Bbb{RP}^1)/G$ where $G$ now acts on the projective line as a subgroup of $\text{SL}(2,\Bbb R)$ rather than $\text{PSL}(2,\Bbb R)$. This I do not understand.

share|cite|improve this question

This all comes down to the lift of the action of $PSL(2,R)$ on $RP^1$ to the 2-fold cover:

The unique connected 2-fold cover of $PSL(2,R)$ is $SL(2,R)$, it acts naturally on the unit circle $S^1$, which is understood as the set of oriented lines in $R^2$. The circle $S^1$ is the unique 2-fold cover of $RP^1$ (defined by forgetting the orientation of the lines). Thus, the group $PSL(2,R)$ lifts to $SL(2,R)$ under the covering map $S^1\to RP^1$. The kernel consists of all $SL(2,R)$ matrices which send each line to itself, i.e. equals $\{\pm I\}$. In particular, the diagram (where the horizontal arrow are group actions and vertical arrow consist of the covering maps and their products) $$ \begin{array}{ccc} SL(2,R)\times S^1 & \longrightarrow & S^1\\ \downarrow &~& \downarrow\\ PSL(2,R)\times RP^1 & \longrightarrow &RP^1 \end{array} $$ is commutative.

The rest is done by applying this consideration to each fiber. Hence, the bundle which is the 2-fold cover Milnor is talking about, is $$ (H\times S^1)/G $$ where $G$ acts on $S^1$ via its isomorphic lift from $PSL(2,R)$ to $SL(2,R)$.

The confusion stems from the fact that $S^1$ is homeomorphic to $RP^1$, but in this situation they should be regarded as different entities.

share|cite|improve this answer

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.