This might be different for sciences, I'm not sure, but mine is a developmental biology PhD...
I'm writing the intro at the mo, and I'm not sure when its ok to cite review papers and when to cite the original papers?
Do you have to have actually read the original papers to cite them? (I mean more than reading the abstract and skimming the paper and then seeing a review and going, oh yeh, i knew that....)

If you want to cite something from a review paper and you haven't accessed the original source then the 'safest' way to do this is to reference as;

blah blah (Jones 1989 cited by Smith 1992)

or

blah blah (1 cited by 2)

Where Jones (or 1) is the original citation given in the review paper and Smith (or 2) is the review paper.

This is 'safest' because you are implying that you haven't read Jones and if Smith got it wrong (and misinterpreted the info from Jones in his/her review) then the blame is on Smith, not on you! Of course, the best option is to get the original paper but this isn't always practical or necessary, especially if it's something peripheral to your main thesis topic.