About 1760 very extraordinary beliefs were current respecting the excessive degree of cold and the rapid variations of temperature which take place in the Arctic regions. Braun, of St. Petersburg, had observed that mercury, in solidifying in the tube of a thermometer, descended through more than four hundred degrees, and it was assumed that the melting point of mercury was about 400 deg. below Fahrenheit's zero. It then became necessary to suppose that, while the mercury in a thermometer was freezing, there was a variation of temperature to this extent, and thus these wild reports became current. Cavendish and Black independently explained the anomaly, and each suggested the same method of determining the freezing point of mercury. Cavendish, however, had a piece of apparatus prepared which he sent to Governor Hutchins, at Albany Fort, Hudson's Bay. It consisted of an outer vessel, in which the mercury was allowed to freeze, but not throughout the whole of its mass, and the bulb of the thermometer was kept immersed in the liquid metal in the interior. In this way the mercury in the thermometer was cooled down to the melting point without commencing to solidify, and the temperature was found to be between 39 deg. and 40 deg. below Fahrenheit's zero. As a chemist, Cavendish is renowned for his eudiometric analysis, whereby he determined the percentage of oxygen in air with an amount of accuracy that would be creditable to a chemist of to-day, and for his discovery of the composition of water; but to the world generally he is perhaps best known by the famous "Cavendish experiment" for determining the mass, and hence the mean density, of the earth. The apparatus was originally suggested by the Rev. John Michell, but was first employed by Cavendish, who thereby determined the mean density of the earth to be 5.45. At the request of the Astronomical Society, the investigation was afterwards taken up by Mr. Francis Baily, who, after much labour, discovered that the principal sources of error were due to radiation of heat, and consequent variation of temperature of parts of the apparatus during the experiment. To minimize the radiation and absorption, he gilded the principal portions of the apparatus and the interior of the case in which it was contained, and his results then became consistent. Cavendish had himself suggested the cause of the discrepancy, but the gilding was proposed by Principal Forbes. As a mean of many hundreds of experiments, Mr. Baily deduced for the mean density of the earth 5.6604. Cavendish's apparatus was a delicate torsion-balance, whereby two leaden balls were supported upon the extremities of a wooden rod, which was suspended by a thin wire. These balls were about two inches in diameter, and the experiment consisted in determining the deflection of the wooden arm by the attraction of two large solid spheres of lead brought very near the balls, and so situated that the attraction of