A detective is now taught that he must use the photograph he is supplied with merely as a check, to make assurance doubly sure, before he ventures on an arrest. What he must principally rely upon is the visual portrait he can evoke in his own imagination, a portrait which, he is told, is only valuable so far as he is able to describe it in words. That which we cannot clearly describe we cannot clearly conceive, is the pith of M. Bertillon's teaching. The pupil is, consequently, made to analyze each feature of the photograph separately, and express the result in certain conventional formulæ that convey a definite meaning to his own mind and to the mind of everybody else who has studied the same method. He makes, in fact, "a portrait in words." The feature that presents the greatest diversity of form and size is the ear, and, strangely enough, the ear is precisely a feature which we hardly ever consciously look at. It has been reserved for M. Bertillon to point out how admirably it is adapted for the purpose of establishing a person's identity. The size of the ear, the relative proportions to one another of the folds, its contour, the surface and shape of the lobe, the manner the lobe is attached to the cheek, and the inclination of the bottom interior ridge known as the antitragus differ most materially in every individual. Let a modern French detective describe an ear as "Deq. cav. vex. tra. sep"; all his colleagues are immediately able to form a mental image of the description of ear he means. [Illustration: THESE ARE THE PORTRAITS OF A CRIMINAL.] [Illustration: THESE ARE PORTRAITS OF AN INNOCENT MAN WHO WAS ARRESTED BY AN UNTRAINED DETECTIVE AS BEING THE SAME MAN, BUT HIS EARS ALONE WERE SUFFICIENT TO ACQUIT HIM.]