LONDON. Royal Society, May 27.—Sir Archibald Geikie, K.C.B., president, in the chair.—Notes concerning tidai oscillatiens upon a rotating globe: Lord Rayleigh.—The absolute value of the mechanicai equivalent of heat in terms of the internationai electrical units: Prof. H. T. Barnes. It is pointed out that the Clark cells used by the author in his determinatiens of the mechanical equivalent of heat in terms of the electrical units were prepared accerding te the old specifications. The absolute measurements of the Clark cell new being carried on with such precision in the various standardising laborateries are expressed in term of the new form of cell with specially prepared mercurous sulphate. There is an impertant difference between the cells, which Wolff and Waters have shown amounts to 0.03 millivolts. The author has compared a set of modem cells with cells set up accerding te the old specification, and finds the same censtant difference. Taking 1.4330 international volts at 15° C. as rcpresenting the modem cells, then the cells made by the old specificatons must to taken Cs 14333 international volts at 15° C. The author's measurements of the mechanical equivalent at different temperatures were calculated on the basis of a value for the Clark cell equal to 1.4342 intemnational volts at 0° C. Re-calculating on the new basis, the value of the mean calorie is found to be 4.1849 joules. This agrees with Reynolds and Meerbys directly determined mean, which, expressed accurately for an interval of temperature between 0° C. und 0° C., comes to 4.1836 joules. Roxviand's mean value between C. and 350 C. 14.185 joules, whiie the auther's value between the seme limits of temperature is 41826 joules. Thus, assuming the variatien of the specific heat of water te ho cerrectlv determined, the value of the Clark cell, equal to 1.4330 intemnational velt, brings the electrically determined mechanical equivalent into excellent agreement with the same censtant measured by mechanical means.—An approximate determination of the boiling points of metals: H. C. Greenwood. Although high temperatures can next to easily attained by means of electric heating, no general investigatien of the boiling points of metals has vet beon carried out. Moreover, such values as aro availabie bave in most cases beon deduced indirectly, and aro very discerdant. In the prosent investigation apparature was devised for directly measuring the temperaturos of ebullition under atmospheric pressure of a censiderable number of metals, alloxving of use up to 2700° C. Heating was effected electrically, and the metal, when unaffected by carbon, was contained in a thin-walled graphite crucible on the outside of which the temperature was estimated by means of a Wanner optical pyremeter. The difference in temperature between the intemnal and extemnal surfaces ef the crucible walls was found te ho negligible. Accuracy of the temperature measurements was secured by checking the pyrometer against the black bedv melting points of specially purifled strips of platinum, rhodium, and iridium. The following values were found: aluminium, 1800° C.; antimony, 1440° C.; bismuth, 1420° C.; chromium, 2200° C.; copper, 2310° C.; iron, 2450° C.; magnesium, 1120° C.; manganese, 1900° C.; silver, 1955° C.; tin, 2270° C.—In dealing with the metals aluminium, chromium, iron, and manganese, which readily combine with carbon, considerable difficulty was experienced in avoiding contact with carbon at the high temperature in questien. This was flnally accomplished by the use of graphite crucibles brasqued with previeusly fused magnesia. In the absence of this protective lining the boiling point was very greatly modified by carburisation. The temperatures indicated for aiuminium and manganese were far belexy those hitherte suppesed necessary for ebullition.