Many attempts have been made in recent months to throw light on the problem of vagueness. That perfect precision is an ideal not to be attained by any language seems clear. But the obvious fact is that words and sentences in our languages are not so precise as we should like to have them, and we are naturally concerned with finding some sort of device by which vagueness can, in the first place, be detected and measured, and, in the second place, eliminated or reduced to a minimum. I shall not be concerned in what follows with these various special attempts to meet the problem. My task is the somewhat more general one of showing that there are other factors contributing to vagueness, one of which is of particular significance when we leave the realm of ostensively defined symbols and begin to concern ourselves with suppositional symbols, constructs and hypotheses.