It is the weight, not numbers of experiments that is to be regarded.