in which the flame is produced. At first separate nipples were employed for the two jets; but, very soon, Mr. Milne hit upon the expedient of drilling two holes, at the required angle, in the same nipple. In this manner, with slight modifications, the burner has continued to be constructed down to the present day. [Illustration: FIG. 3.--FISHTAIL BURNER.] The explanation of the preference accorded to this burner over its predecessor, the batswing, is to be found chiefly, I think, in the very different shapes of the respective flames produced by the two burners. The batswing, in its original form, produced a flame of great width, but of no corresponding height. The extremities of the flame, stretching out from the burner so far on either hand, were easily affected by an agitation of, or commotion in the surrounding atmosphere; a slight draught or current of air causing the flame to smoke at these points. The extreme width of flame also precluded the use of this burner in globes. The flame produced by the union-jet burner, as first constructed, was very different to the one just described. Longer than that of the batswing, and considerably narrower (but widening gradually from its base, at the burner, to its apex), it presented somewhat nearly the appearance of an isosceles triangle; or more closely, perhaps (with its slightly-forked apex), the tail of a fish, from which resemblance it is commonly designated the fishtail burner. This form of flame was better adapted for use in globes, and also better withstood the effects of draughts. And it is perhaps not unreasonable to suppose that as in shape it approached more closely to the kind of flame with which the people had been familiar in oil lamps, the flame produced by the union-jet burner was more agreeable to the eye than that of the batswing, and that this seemingly trivial consideration will account, to some extent at least, for the undue favour shown towards it. For it must not be assumed, because of the widespread popularity to which the union-jet so early attained, and which it has continued to enjoy, that it was of necessity a better burner (in the sense of developing more light for the gas consumed) than the one which preceded it. On the contrary, in this regard it was not quite so effective as the batswing. Nor is this result surprising, looking at the different methods adopted in the two burners for producing the same effects of light and flame. [Sidenote: Union-jet and batswing burners compared.]