"The only positive finding which could be drawn from the first series, was the conclusion that the relationships obviously had a more complicated lay-out than had been thought, for the effects were so varied that no obedience to any law could be discovered."