Recalculations on earlier experiments on the fountain effect below 1°K