A function of two variables is said to be homogeneous of degree n if there is a constant n such that $$f(tx,ty)=t^nf(x,y)$$ for all $t, x, y$ for which both sides are defined.
A differential equation of the form
$$M(x,y)dx+N(x,y)dy=0$$
is homogeneous if $M$ and $ N$ are homogeneous of the same degree. You can solve homogeneous equations by turning them into separable equations using the substitution $$y=vx.$$
My question is, why is this substitution mathematically valid? I understand that for every $y$ and $x$ there exists a $v$ such that $y=vx$, but using this substitution seems to suggest that there exists a $v$ such that $y=vx$ for every $x$, or that you're just avoiding dealing with the differentials in $x$ by inventing a variable $v$ and varying that instead of $x$ or something.
I understand how to proceed algorithmically once the substitution is made, and how, if you assume it all makes sense, that it can lead to nice tidy solutions, but I don't really understand--analytically, I guess?--what we're actually doing by making this substitution.