why does adjusted R-squared gets smaller?

[bibimiwi]

Hello,
I have got a question concerning the adjusted R-squared.
I estimated an OLS model using X1 X2 D1 D2 as explanatory variables. D1 and D2 are dummy variables. For this model I get an adjusted R-squared of 0.13.
Then I wanted to specify my model and introduced X1D1 and X2D2. Now my adjusted R-squared is reduced to 0.11. What does it get smaller? Should it not always increase the more explanatory variabled I use, as it is the case with R-squared?

Thank you very much

[EViews Gareth]

http://en.wikipedia.org/wiki/Coefficien ... djusted_R2