“When we wish to make a great many observations, it is very convenient to mark divisions on the table (which carries the lights), in order to read off, by means of them, the distance of the lamps from the shadows which they illuminate. By this means, each observation need not occupy more than two minutes. I generally use a table CC DD (fig. 97), about two mètres long (6 feet 6 inches), by 80 centimètres wide (2 feet 8 inches). At one end I place the screen AB, covered with white paper, dull (or not glazed), and kept in a vertical plane by two small pieces P and Q. Through the point M, the centre of the opaque body, I draw two lines M _f_ and M _g_, equally inclined to the central line _x_ _y_, whose extremities _b′_, _a′_ are the axes of the two shadows. These lines must be inclined in such a manner that the distance of the shadows may be a little less than the diameter of the opaque body, or so that they may actually touch each other, according to the mode of observing which you wish to follow. These lines M _f_, M _g_ I divide into decimètres and centimètres, starting from the points _a′_, _b′_ and over these lines I place the centres of the flames; the distance between the shadows remains always the same, whatever may be the distance of the lamps: to determine the distance of each lamp from the shadow which it illuminates, we ought, strictly speaking, to take the distance of the centre of the flame _b_ from the point _a′_; but as the distance from the point _b_ to the point _a′_ differs little from the distance between the points _b_ and _b′_, we assume the latter for the former, without causing any sensible error. That distance may be obtained very conveniently by taking the half of the sum of the distances of the two extremities _z_ and _z′_ of the diameter of the pedestal of the lamp. When the burner is not placed over the centre of the pedestal, we may suspend from it a small plummet, whose point will touch some division and indicate the distance between the centre of the burner and the shadow. “When the lights are coloured, the shadows are coloured also, and it is then far more difficult to judge accurately of their intensity. They may in that case be much better seen from the point _x_, as the black opaque body which is interposed between them renders the difference of colour less sensible to the eye. “The opaque body M is a cylindric rod of iron, whose upper part is blackened in the flame of a lamp, in order to prevent the reflection which might interfere with the _sharpness_ (_netteté_) of the shadows, and to make them more distinct when they are viewed from the point x.”[86] [86] Those who feel a curiosity to look farther into this subject may consult Count Rumford’s elaborate paper in the Phil. Trans. for 1794, p. 67.