According to dispositionalism about modality, a proposition \u3cp\u3e is possible just in case something has, or some things have, a power or disposition for its truth; and \u3cp\u3e is necessary just in case nothing has a power for its falsity. But are there enough powers to go around? In Yates (2015) I argued that in the case of mathematical truths such as \u3c2+2=4\u3e, nothing has the power to bring about their falsity or their truth, which means they come out both necessary and not possible. Combining this with axiom (T), it is easy to derive a contradiction. I suggested that dispositionalists ought to retreat a little and say that \u3cp\u3e is possible just in case either p, or there is a power to bring it about that p, grounding the ...