>>33

So this was right.

>>>/cbts/54610

>>>/cbts/54658

Assuming the gaps have something to do with size of the squares increasing although not quite gotten this yet