Great, thank you Christian!

]]>Dear Dino, this is exactly what this appendix states (and I have no reason to doubt it as I was in the process of re-establishing it when I got pointed out to it).

]]>In my work, in each iteration I would like to use a normal proposal density for a parameter x_1 where the proposal mean and covariance matrix depend on the current values of parameters x_2 and x_3 which depend on the previous value of x_1 and so on. While this works well in practice, I was wondering if such a scheme preserves ergodicity (it seems it does!) and if to treat the proposal density as independent from the previous value of x_1 when calculating the acceptance ratio. Would highly appreciate any thoughts on this. ]]>