The annexed figure shows the form adopted by the inventor for oil and kerosene lamps. As may be seen, the chimney consists of two cones, A and B, connected end to end by their small bases. The upper one, A, or divergent cone, is constructed according to a variable angle, but one which, in order to produce its maximum effect, ought not to differ much from 5°. This cone rests upon the convergent one, B, whose angle, as we have said, varies between 35° and 45°. To the large base of this cone there is soldered a cylindrical part, c, designed for fixing the chimney to the holder. The height given the divergent cone is likewise variable, but a very beautiful light is obtained, when it is equal to six times the diameter of the contracted part. When the lamp is designed to be used in a still atmosphere, free from abrupt currents of air, the height may be reduced to four times the diameter of the base, without the light being thereby rendered any the less bright. As for the height to be given the convergent cone, B, that is determined by the opening of the angle according to which it has been constructed. Finally, as a general thing, the diameter of the small base should be equal to half the large base of the convergent cone, B. The new chimney should be placed upon the holder in such a way that the upper part of the wick tube, D, is a few millimeters beneath the base of the convergent cone. The height to be given the wick varies according to the lamp used. It is regulated so as to obtain a steady and regular combustion. In oil lamps it must project about 1½ centimeters. If two lamps of the same size be observed, one of which is fitted with the new chimney and the other with the old style, we shall be struck with the difference that exists in the color of the flame as well as in its intensity. While in the case of the cylindrical glass the flame is red and dull, in that of the circuit it is white and very bright. This, however, is not surprising when we reflect upon the theoretical conditions upon which the construction of the new chimney is based--the strong influx of air having the result of causing a more active combustion of the liquid, and consequently of raising to white heat the particles of carbon disseminated through the flame. As it was of interest to ascertain what the increase of illuminating power was in a given lamp provided with the new chimney, Mr. Felix le Blanc undertook some photometric experiments. The trials were made with a Gagneau lamp provided with a chimney of the ordinary shape, and then with one of Mr. Bayle's. The measurements were made after each had been burned half an hour. The light of the standard Carcel lamp being 1, there was obtained with the Gagneau lamp with the ordinary chimney 1.113 carcels, and with the Bayle chimney 1.404 carcels. Thus 1.113:1.404 represents the ratio of the same lamp with the ordinary chimney and with that of Bayle. Whence it follows that the light of the