Most specifically, his use of small epsilon introduced at the end of section 1 of "Types, Abstraction and Parametric Polymorphism" is throwing me, but in general I would like references to symbols in the Type and Logical Relations literature.
You aren't going to find a list of all notations ever used. Apart from a few standard notations, each paper will define the notation that it uses.
This paper is from the dark ages of typewriter typesetting, so the typographic quality is subpar. The character
ε stands for $\in$, which is the standard mathematical notation for set membership.