1: \begin{abstract}

2: We consider parabolic stochastic partial differential equations driven

3: by white noise in time. We prove exponential convergence of the transition probabilities towards a unique invariant measure under suitable conditions. These conditions amount essentially to the fact that the equation transmits the noise to all its determining modes. Several examples are investigated, including some where the noise does {\it not} act on every determining mode directly.

4: \end{abstract}

5: