Experimental science hardly ever affords us more than approximations to the truth; and whenever many agents are concerned we are in great danger of being mistaken.