The key point is that the Sylow $q$-subgroups of $G/P$ (for any prime $q$) are precisely the homomorphic images of the Sylow $q$-subgroups of $G$ under the canonical homomorphism $G\to G/P$. Also, homomorphic images of abelian groups are always abelian. E.g., this is elaborated in both Pete's and Boris' excellent answers. I've provided details in my answer below.
Firstly, let's consider the case of primes $q\neq p$: any Sylow $q$-subgroup of $G/P$ (in fact, any subgroup of $G/P$) is of the form $H/P$ for some subgroup $H$ of $G$ containing $P$. Now, $H=QP$ where $Q$ is a Sylow $q$-subgroup of $H$. (Why?) Since $Q$ is also a Sylow $q$-subgroup of $G$ (why?), it follows that $H/P$ equals the image of $Q$ under the canonical homomorphism $G\to G/P$.
Now let's consider the case $q=p$. If a $p$-subgroup $P\subseteq G$ is normal in $G$, then $P$ must be contained in all Sylow $p$-subgroups of $G$. (Exercise; use the Sylow theorems.) The intersection of all Sylow $p$-subgroups of $G$ is referred to as the $p$-core of $G$ and is the unique maximal normal $p$-subgroup of $G$.
Therefore, the Sylow $p$-subgroups of $G/P$ are precisely the Sylow $p$-subgroups of $G$ which contain $P$.
I hope this helps!