per or per  ?

edit

The article declares that the ielative intensity noise is measured in units  . The only reference (shot noise) suggests, that fluxtuations are proportional to square of frequency band; so, the relative amplitude of fluctuations sohuld have dimension  . Should not we specify that the REM (Relatice Intensity Noise) is proportional to the mean square of the amplitude of relative fluctuations? May be not only "proportional to", but just "equal to"? dima 01:33, 4 April 2007 (UTC)Reply