1: \begin{abstract}

2:   In this paper the numerical solution of non-autonomous

3:   semilinear stochastic evolution equations driven by an additive

4:   Wiener noise is investigated. We introduce a novel fully discrete

5:   numerical approximation that combines a standard Galerkin finite element

6:   method with a randomized Runge--Kutta scheme. 

7:   Convergence of the method to the mild solution is proven with

8:   respect to the $L^p$-norm, $p \in [2,\infty)$.

9:   We obtain the same temporal order of convergence as for

10:   Milstein--Galerkin finite element methods but without imposing any

11:   differentiability condition on the nonlinearity. The results are 

12:   extended to also incorporate a spectral

13:   approximation of the driving Wiener process. An application to a stochastic

14:   partial differential equation is discussed and illustrated through a

15:   numerical experiment. 

16: \end{abstract}

17: