as a flux on the entire sand heated in the usual manner. The sand so treated is reddened in the first instance, changes afterwards to an opaque white, and finally is very slightly agglutinated, resembling in its shades of colour and in its cohesion that which adheres to the outermost coating of the tube.
The sand was exposed to the flame of a spirit lamp, urged by a stream of oxygen gas, in which flame Dr. Marcet has shewn that a thick wire of platina may be melted. The grains of hornstone porphyry first began to flow, and presently acting upon the grains of quartz, formed a clear glass mingled with portions of an olive colour. Even thus however the fusion was partial, and the utmost intensity of the flame was required to support it.
The glass thus formed resembled that of the tube in its external characters, in being very hard and difficult of fusion, and in containing a great excess of siliceous matter: for it was ascertained that a fragment of the tube was scarcely softened at the edges by the same flame that had melted the sand; and that the substance of the tube contained a large portion of silica, had been determined by one of the members.
From what has been stated, it appears that the tubes have all the marks of fusion, and that their substance can be imitated in some measure by subjecting the sand to intense heat. That they are of very recent date is certain from the shifting nature of the sand hillocks in which they are found, and from their inability to remain alone and unsupported by the sand without breaking. Lightning seems to be the only agent that could at once supply the heat and the force requisite to make them. In the familiar experiment of perforating a quire of paper by the Leyden battery, a mechanical effect of electricity analogous to the present is exhibited, and instances of fusion by the same instrument it is not necessary to enumerate.
3 x 2