diffhistmSpectral theorem22:28−30Henrik Schumachertalkcontribs(Of course, Dirac delta distributions lie in some Banach space, e.g., in that dual space of the Banach space of continuous functions. The reference to Banach spaces here was a read hering. As a side note: the elements of Lebesgues space are no "classical functions", either, but rather _equivalence classes_ of functions. But I think the remainder of the statement can be left as is, nonetheless.)