In the proof given by the book of Shafarevich "Basic algebraic geometry 1", section 5.3 of chapter 1, that every finite map between algebraic varieties $f:X \longrightarrow Y$ (where we assume $f(X)$ is dense in $Y$ and finite means that $k[X]$ is an integral extension of $k[Y]$), he uses that the restriction of $f$ to a closed set $Z$, $f':Z \longrightarrow \overline{f(Z)}$ is also finite. Why is this so?
Thanks in advance
Algebraically, this is the fact that if $\varphi: R \to S$ is an integral extension, then $R/\varphi^{-1}(I) \to S/I$ is again integral. This is more or less immediate to prove -- if $S$ is finitely-generated as an $R$-module, then so is $S/I$, by the same generators. And the same is true whether you think of $S/I$ as an $R$- or $R/\varphi^{-1}(I)$-module. (Here $R = k[Y], S = k[X]$ and $I = I(Z)$ in $X$.)
Geometrically, this is essentially the fact that the following maps are all finite:
So $Z \hookrightarrow X \to Y$ is finite; and so is the "restriction" of $Z \to Y$ to $\overline{f(Z)}$.