I'm sure I recall Quine saying in various places that one distinction between logic and set theory is that logic should have no ontological presuppositions (or, at most should presuppose some thing exists). He followed a line that Russell also took, from 1903 through Principia Mathematica, that existence claims such as the axiom of infinity are not logical truths. And I find on the Stanford Encyclopedia that one reason Quine gave to calling second order logic set theory, as opposed to first order logic which he considered properly logic, is "that first-order logic has no ontological presuppositions of its own." But no citation is given.
Where does Quine say things like this?