The chimney is not only an apparatus designed to carry off the smoke and gases due to combustion, for its principal role is to break the equilibrium of the atmospheric air, which is the great reservoir of oxygen, and to suck into the flame, through the difference of densities, this indispensable agent to combustion. The lamps which we now use are provided with cylindrical chimneys either with or without a shoulder at the base. The shouldered chimney would be sufficient to suck in the quantity of air necessary for a good combustion if we could at will increase its dimensions in the direction of the diameter or height. But, on account of the fragile nature of the material of which it consists, as also because of the arrangement of the lighting apparatus, we are forced lo give the chimney limited dimensions. The result is an insufficient draught, and consequently an imperfect combustion. It became a question, then, of finding a chimney which, with small dimensions, should have great suctional power. Mr. Bayle has taken advantage of the properties of convergent-divergent ajutages, and of the discovery of Mr. Romilly that a current of gas directed into the axis and toward the small base of a truncated cone, at a definite distance therefrom, has the property of drawing along with it a quantity of air nearly double that which this same current could carry along if it were directed toward a cylinder. In getting up his new chimney, Mr. Bayle has utilized these principles as follows: Round-burner lamps have, as well known, two currents of air--an internal current which traverses the small tube that carries the wick, and an external one which passes under the chimney-holder externally to the wick. In giving the upper part of the chimney, properly so called, the form of a truncated cone whose smaller base is turned toward the internal current of air, that is to say, in directing this current toward the contracted part of the upper cone, at the point where the depression is greatest, a strong suction is brought about, which has the effect of carrying along the air between the wick and glass, and giving it its own velocity. The draught of the two currents having been effected through the conical form of the upper part of the chimney, it remained to regulate the entrance of the external current into the flame. If this current should enter the latter at too sharp an angle, it would carry it toward the mouth of the chimney before the chemical combustion of the carbon and oxygen was finished; and if, on the contrary, it should traverse it at too obtuse an angle, it would depress and contract it. Experience has shown that in the majority of cases the most favorable angle at which the external current of air can be led into the flame varies between 35° and 45°. We say in the majority of cases, for there are exceptions; this depends upon the combustive materials and upon the conditions under which they enter the flame.