Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Exact sequence & meaning of a corollary?

  1. Apr 21, 2007 #1
    on page 4 of the March 17th lecture found at
    (you need to scroll down a bit to see the link the March17

    In the first corollary, when they say map 1 = map 4 = 0 do
    they mean these are 0-homomorphisms?

    Another other dumb question I had is about the bottom
    corollary on the same page. What is it saying/what does it
    mean? Where could I find a sketch of it's proof? That
    might help me better understand the corollary. Or is
    an accurate sketch of its proof basically be the sequence of
    theorems & corollaries above it?

    Thanks a lot :)
    Last edited by a moderator: Apr 22, 2017
  2. jcsd
  3. Apr 22, 2007 #2

    Chris Hillman

    User Avatar
    Science Advisor


    Can I offer a bit of advice about asking questions like this? I think you should get in the habit of attributing a writing to an author by name, not by pronoun ("they"). Like bad grammar or bad spelling, this can make a poor impression on at least some of the scholarly minded PF members who are best able to help you. More to the point, there are good reasons for the scholarly convention of citing your sources carefully, and these reasons apply to questions posted in forums like PF, just as they would in a email to a faculty member at a university, or in a research paper. Ultimately it is in your own best interest, as well as that of lurkers who might have similar questions, to observe such basic scholarly conventions to the best of your ability. IOW, when in Rome, do as the romans do. Or in still other words, the point of the motto of William F. Friedman (quoted in the title of my reply) is that taking the time to think things through usually pays off in the end.

    As as example, it would have been most helpful had you stated that you are reading

    Last edited by a moderator: Apr 22, 2017
  4. Apr 22, 2007 #3
    I hadn't been aware of how to properly post/compose messages in these sorts of forums. So thank you for the advice. I now know.

    No I am not a student of Prof. Tolman. I found her lecture notes online, the notes say they are from 2002.

    I meant the very last corollary on page 4. "Assume all the fixed points ...". I had noticed the last page had a summary of the notation so I am aware of what all the symbols mean. To attempt to make my question more specific, is there a geometrical or simplified description/statement of what this corollary is saying? I am having difficulty seeing the big picture w.r.t. this corollary, despite already having studied the earlier portions.
  5. Apr 23, 2007 #4

    Chris Hillman

    User Avatar
    Science Advisor

    Did you try to email Prof. Tolman?
  6. Apr 23, 2007 #5
    No I did not. I do not know Prof. Tolman, nor do I go to UIUC. I know that is no excuse. I merely thought I would first try to find somebody that might be able to help me understand this math on a forum like this one.
  7. Apr 24, 2007 #6


    User Avatar
    Science Advisor
    Homework Helper

    the answer to the first question is yes. although i do not see how she deduces all that from that hypothesis.

    ok i guess the star at the top means the maps 1 and 4 really are equal, so since 1 =0 then also 4=0. it then follows immediately from the definition of the word exact that 2,3 re as stated.

    as to the emaning of the corollary, you have to read the notation on oprevious pages.

    then you see she is trying to calculate the "equivariant cohomology" of the action by S^1. and she is saying how ti find generatiors by looking only at the fixed points.

    this is a bsic principle of such actions, everythig is eklarned from looking at the fixed points.

    more precisely she is telling you everything in the equivariant cohomology is coming from the euler class of the bundle E in the following way:

    each fixed point ahs a certaib height measured by the morse fucntion phi.

    for each fixed point p, there is an equivariant cohomology class alpha(p), that equals the eukler clas at p but that vanishes at fixed points below p.

    moreover these generate. so she is decomposing the eukler class into sort of summands, one for each fixed point, and doing so in order of their height on the manifold.

    does this help?
    Last edited: Apr 24, 2007
  8. May 1, 2007 #7
    Yes, thank you! It did help :) I finally understand the last few examples now too thanks to your explanations.

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook