# Talk:Banach manifold

WikiProject Mathematics (Rated Start-class, Mid-importance)
This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of Mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Mathematics rating:
 Start Class
 Mid Importance
Field:  Geometry

can someone please explain, what is 'an r-times continuously differentiable function'? How is differentiability for mappings between banach spaces defined? or is this somehow inplicit in the abstract tangent space notion?

It's a little implicit, but not too abstract, and is formally just the same as Fréchet differentiability for maps between finite-dimensional real spaces. A map f : U → Y defined on an open subset U of a Banach space X and taking values in a second Banach space Y is differentiable if, for each point x of U, there is a linear map (the derivative (df)(x)) from TxU = X into Y such that [insert your favourite version of Fréchet derivative here]. Put another way, the derivative df is a function U → Lin(XY). But Lin(XY) is itself a Banach space, so if df : U → Lin(XY) is smooth enough, we can rinse and repeat to get a second derivative d2f : U → Lin(X; Lin(XY)). Up to isomorphism, Lin(X; Lin(XY)) is the same as Lin(X × XY). In other words, at each point x of U, (d2f)(x) is a bilinear map, one that "eats" two tangent directions to U at x and "spits out" a tangent vector to f(U) at f(x). Continuing in this fashion, the rth derivative drf is a map from U into Lin(XrY). Just as in elementary analysis, an r-times continuously differentiable function is one for which drf exists and is a continuous function. Sullivan.t.j 00:56, 2 July 2007 (UTC)

## A Global Chart ?

I would like to add an important fact into this article, but I'm not sure if this is really true: "Every infinite dimensional Banach manifold admits a global chart, i.e. every infinite dimensional Banach manifold is diffeomorphic to a subset of Banach space."

Does anyone know if this statement is correct? If so, it would a good thing to add it because it points out one main difference between the (inf.dim.) Banach case and the ordinary finite dimensional case. (in finite dimensions it remains true that there is always a finite atlas - which is a surprising fact for its own) --131.234.106.197 (talk) 11:59, 16 April 2008 (UTC)

This is an interesting question, but I think that it first needs a little refinement: is it true that every infinite-dimensional Banach manifold, M, is diffeomorphic to some open subset of some Banach space, Y? Sullivan.t.j (talk) 15:17, 16 April 2008 (UTC)
It turns out that the result is true if the space on which the manifold is modelled is separable. I will add the theorem and a reference forthwith. Sullivan.t.j (talk) 17:24, 16 April 2008 (UTC)

## Math priority rating

I think that this concept should receive at least a "mid importance rating" in comparison to other articles since "low importance" is simply too low in my opinion. Any comments? --PST 07:52, 9 September 2009 (UTC)

Mid seems reasonable to me. Sławomir Biały (talk) 00:34, 11 September 2009 (UTC)
Thanks, I have changed the importance rating. --PST 04:41, 15 September 2009 (UTC)