From the cockspur and single jet burners the gas ascended in streams, rising into the air until it came in contact with sufficient oxygen to completely consume it. In order that this might take place without producing a flame of an inordinate length, and without much smoke, the orifices were restricted to a very small size; and the gas issuing from these at considerable pressure tended to draw in, and mix with the air in its course. Besides the loss of illuminating power caused by this mixture of air with the gas flame (similar to what takes place in a Bunsen burner), the cooling influence upon the small body of flame of the mass of metal composing the burner, operated still further to reduce the quantity of light which the gas was calculated to yield. With the batswing the gas was spread out producing, when ignited, a thin sheet of flame, by which means the gas was enabled to combine more readily with the air necessary to effect complete combustion. The size of the flame being, in comparison with that of the cockspur, so much larger proportionately to the metal burner, the cooling effect of the latter was not so apparent. The increased size of flame, also, of itself, tended to improve the illuminating power; each portion of flame contributing to elevate and sustain the temperature of the whole, and so to heighten the intensity of incandescence to which the light-giving particles were raised. [Sidenote: Batswing and Argand burners compared.] Even with the Argands of that day, the batswing compared not unfavourably. The former burner, having the regulation of its air supply under complete control, gives the best results when the gas is supplied to it at a low pressure; as then the requisite quantity of air to ensure complete combustion of the gas can be delicately adjusted by means of a chimney of suitable length. When the gas and air have been nicely adjusted to each other, the flame becomes extremely sensitive to any change of pressure in the gas supply; a diminution of the supply, by reducing the quantity of gas issuing from the burner without at the same time proportionately diminishing the supply of air, tends to destroy the illuminating power by the cooling action of the surplus air; while an increased pressure, by allowing more gas to issue than the air can consume, causes the flame to smoke. But at the time to which I now refer the principles of combustion were little understood, still less applied in the construction of burners. Besides this, the pressure of the gas in the mains was excessive; and there being no method adopted of controlling it at the burner, the construction of a good Argand was, under the circumstances, almost impossible. The batswing was not so prejudicially affected by an excess of pressure. Pressure to some extent was, indeed, required to enable the flame to attain its normal shape; while any excess forced the gas through the flame without permitting it to be raised to incandescence before being