I've just had a look at the EV blog on 15squared - it has MEM in the empty cells and I expect the printed solution in the 'i' will too.
Well done to those (the majority, I think) who got it right - not me though!
I hope the 'alternative' version at least gets a mention over there...