Persons and society must, then, be studied primarily in the imagination. It is surely true, _prima facie_, that the best way of observing things is that which is most direct; and I do not see how anyone can hold that we know persons directly except as imaginative ideas in the mind. These are perhaps the most vivid things in our experience, and as observable as anything else, though it is a kind of observation in which accuracy has not been systematically cultivated. The observation of the physical aspects, however important, is for social purposes quite subsidiary: there is no way of weighing or measuring men which throws more than a very dim side-light on their personality. The physical factors most significant are those elusive traits of expression already discussed, and in the observation and interpretation of these physical science is only indirectly helpful. What, for instance, could the most elaborate knowledge of his weights and measures, including the anatomy of his brain, tell us of the character of Napoleon? Not enough, I take it, to distinguish him with certainty from an imbecile. Our real knowledge of him is derived from reports of his conversation and manner, from his legislation and military dispositions, from the impression made upon those about him and by them communicated to us, from his portraits and the like; all serving as aids to the imagination in forming a system that we call by his name. I by no means aim to discredit the study of man or of society with the aid of physical measurements, such as those of psychological laboratories; but I think that these methods are indirect and ancillary in their nature and are most useful when employed in connection with a trained imagination.