Thanks Ich,
It took a bit of working out. I originally started applying the approximation you gave to the WHOLE of the LHS of the first line, but I ended up with a half of Tau in the result. I then realized I just apply it the the second argument of the LHS and again on the right (after a bit...