It has been recognized that Peirce's gamma part of his Existential Graphs (EG) contains that kind of treatment of modal logic that is today called model-theoretic. Peirce is even supposed to have anticipated possible-worlds-semantics (PWS).¹ In this paper I want to explore how close he got to our contemporary semantics for modal logics and if the differences one finds can only be seen as evidence of an archaic and outdated understanding of logical matters or if they have a hidden potential that might be fruitfully applied in the discussions of problems within the now dominant interpretation of logic.