\magnification 1800 \parskip 25pt \parindent 0pt \hoffset -0.3truein \hsize 7truein \def\N{{\bf N}} \def\S{\Sigma_{i=1}^n} \def\R{{\bf R}} \def\C{{\bf C}} \def\x{{\bf x}} \def\y{{\bf y}} \def\e{{\bf e}} \def\i{{\bf i}} \def\j{{\bf j}} \def\k{{\bf k}} \def\ep{\epsilon} \def\f{\vskip 10pt} \def\u{\vskip -8pt} \def\emph{{\it}} \def\textit{{\it}} Defn: $(\varphi, U)$ is a \emph{\it chart } or {\it coordinate neighborhood} if $\varphi:U\rightarrow U^{\prime}$ is a homeomorphism, where $U$ is open in $M$ and $U^{\prime}$ is open in $\R^n$. Defn: Two charts, $(\varphi, U)$ and $(\psi, V)$ are {\it $C^{\infty}$ compatible} if the function $\varphi\psi^{-1}: \psi(U\cap V)\rightarrow\varphi(U\cap V)$ is a diffeomorphism. Defn: A \emph{\it (pre) atlas } or {\it differentiable or smooth structure} on $M$ is a collection of charts on $M$ satisfying the following two conditions: \u\u \hskip 30pt i.) the domains of the charts form an open cover of $M$ \u\u \hskip 30pt ii.) Each pair of charts in the atlas is compatible. Defn: An atlas is a (maximal or complete) atlas if it is maximal with respect to properties i) and ii). A {\it differential (or smooth or $C^{\infty}$) n-manifold} \emph{M} is a topological n-manifold together with an atlas. Theorem: A (pre) atlas can be uniquely enlarged to a maximal atlas. \end \eject Thm: $f$ is differentiable at {\bf a} implies $f$ is continuous at {\bf a}. \vfill Thm: Let $f: \R^n \rightarrow \R^m$, $f = (f_1, ..., f_m)$. $f$ is differentiable at {\bf a} iff $f_i: \R^n \rightarrow \R$ is differentiable at {\bf a} for all $i = 1, ..., m$ Thm: If $f$ is differentiable at {\bf a} then ${\partial f_i \over \partial x_j}$ exists for all $i, j$ and $Df({\bf a})$ = the Jacobian evaluated at {\bf a}. Thm: Let $f: \R^n \rightarrow \R^m$, $f = (f_1, ..., f_m)$. If ${\partial f_i \over \partial x_j}$ exists and are continuous in a neighborhood of {\bf a} for all $i, j$, then $f$ is differentiable at {\bf a} Ex: Is $f(x, y) = x^2y$ differentiable at (3, 1). \vskip 20pt Find the equation of the tangent plane to $f(x, y) = x^2y$ at (3, 1). Estimate f(3.1, .9) \eject 2.4 Thm: If $f, g: \R^n \rightarrow \R^m$ is differentiable at {\bf a}, then $f + g$ is differentiable at {\bf a} and $D(f + g) = Df + Dg$. Thm: Let $c \in \R$. If $f: \R^n \rightarrow \R^m$ is differentiable at {\bf a}, then $cf$ is differentiable at {\bf a} and $D(cf) = cDf$. Thm: If $g: \R^n \rightarrow \R^m$ is differentiable at {\bf a} and if $f: \R^m \rightarrow \R^k$ is differentiable at g({\bf a}), then $f \circ g$ is differentiable at {\bf a} and $D(f\circ g)({\bf a} ) = Df(g({\bf a}) ) Dg({\bf a} )$. Note for the product and quotient rule, $f, g$ are real-valued functions, NOT vector valued. Thm: If $f, g: \R^n \rightarrow \R$ is differentiable at {\bf a}, then $fg$ is differentiable at {\bf a} and $D(fg) = g( {\bf a})Df( {\bf a}) + f( {\bf a})Dg ( {\bf a})$. Thm: If $g({\bf a}) \not= 0$ and $f, g: \R^n \rightarrow \R$ is differentiable at {\bf a}, then $f/g$ is differentiable at {\bf a} and $D(f/g) = {g( {\bf a})Df( {\bf a}) - f( {\bf a})Dg ( {\bf a}) \over g({\bf a})^2}$. \end