1: \begin{abstract}In an indirect Gaussian sequence space model lower and

2:   upper bounds are derived for the concentration rate of the posterior

3:   distribution of the parameter of interest shrinking to the parameter

4:   value $\TrSo$ that generates the data. While this establishes posterior

5:   consistency, however, the concentration rate depends on both $\TrSo$

6:   and a tuning parameter which enters the prior distribution. We first

7:   provide an oracle optimal choice of the tuning parameter, i.e.,

8:   optimized for each $\TrSo$ separately. The optimal choice of the

9:   prior distribution allows us to derive an oracle optimal

10:   concentration rate of the associated posterior distribution.

11: Moreover, for a given class of parameters and a suitable choice of the

12: tuning parameter, we show that the resulting uniform

13:   concentration rate over the given class is optimal in a minimax

14:   sense.  Finally, we construct a hierarchical

15:   prior 

16:   that is adaptive. This means that, given a parameter $\TrSo$ or a

17:   class of parameters, respectively, the posterior distribution

18:   contracts at the oracle rate or at the minimax rate over the

19:   class. Notably, the hierarchical prior does not depend neither on $\TrSo$ nor

20:   on the given class. Moreover, convergence of the fully data-driven Bayes

21:   estimator at the oracle or at the minimax rate is  established.

22: \end{abstract}

23: