porcelain-clay is refractory in the highest temperatures that we have it in our power to produce in furnaces. Reaumur made another curious observation on glass, which has been, since his time, employed very successfully to explain the appearances of many of our trap-rocks. If a glass vessel, properly secured in sand, be raised to a red heat, and then allowed to cool very slowly, it puts off the appearance of glass and assumes that of stoneware, or porcelain. Vessels thus altered have received the name of _Reaumur’s porcelain_. They are much more refractory than glass, and therefore may be exposed to a pretty strong red heat without any danger of softening or losing their shape. This change is occasioned by the glass being kept long in a soft state: the various substances of which it is composed are at liberty to exercise their affinities and to crystallize. This makes the vessel lose its glassy structure altogether. In like manner it was found by Sir James Hall and Mr. Gregory Watt, that when common greenstone was heated sufficiently, and then rapidly cooled, it melted and concreted into a glass; but if after having been melted it was allowed to cool exceedingly slowly, the constituents again crystallized and arranged themselves as at first--so that a true greenstone was again formed. In the same way lavas from a volcano either assume the appearance of slag or of stone, according as they have cooled rapidly or slowly. Many of the lavas from Vesuvius cannot be distinguished from our _greenstones_. Reaumur’s labours upon the thermometer must not be omitted here; because he gave his name to a thermometer, which was long used in France and in other parts of Europe. The first person that brought thermometers into a state capable of being compared with each other was Sir Isaac Newton, in a paper published in the Philosophical Transactions for 1701. Fahrenheit, of Amsterdam, was the first person that put Newton’s method in practice, by fixing two points on his scale, the freezing-water point and the boiling-water point, and dividing the interval between them into one hundred and eighty degrees.