240
edits
(typo) 

This argument is clarified by taking the limit carefully. If an approximate charge operator is applied to the vacuum,
:<math>
{d\over dt} Q_A = {d\over dt} \int_x e^{x^2\over 2A^2} J^0(x) = \int_x e^{x^2\over 2A^2} \nabla \cdot J = \int_x \nabla(e^{x^2\over A^2}) \cdot J
</math>

edits