# Integration by parts with ill-behaved functions.

1. Sep 30, 2011

### Tomer

This is a general question: as far as I know, integration by parts is allowed only with functions that are continuously differential.
However, I'm reading Griffiths Quantum book, and he easily uses this technique in integrals involving the delta "function" and the step "function", without trying to justify it.
Is there some sort of an extension to it? It's all peculiar since the delta function isn't a function anyway, so the whole definitions kinda lose their strictness, but without delving into it's mathematical definition (I'm good so far with the integral relations it holds), is there a way to justify integration by parts?

Thanks,
Tomer