TRANSACTIONS OF SECTION B. 611 



These results are plotted in the figure. 



If Cj,2 be the mean specific heat between two temperatures t^ and t„, we have 

 the relation — 



C,,3(^3 - O = C,.,(i, - t,) + C,;{t,-t,-). 



If then we take the mean specific heat from 20° C. 



to 470° to be 0-199 



„ 750° „ 0-21G 



„ 880° „ 0-278 



and „ 1,190° „ 0-290, 



we get the following results : — 



The mean specific heat between — 



20° and 470° is 0'199 

 470° „ 750° „ -243 

 750° „ 880° „ -62G 

 880° „ 1,190° „ -323. 



As Mr. Fisher was anxious to use our results in some calculations, we supplied 

 him with approximate numbers before all our observations were completed. They 

 do not, however, differ much from the above. 



The general result seems to be that the specific heat of basalt follows the 

 ordinary rule that the specific heat of a substance is greater in the liquid than in 

 the solid state. There is a large absorption of heat in the neighbourhood of 800°, 

 which raises the mean specific heat between 750° and 880° to the large value 

 of 0-626. 



An Apparatus for Testuig Safefy Lamps. 

 By Professor F. Clowes, F.G.S. 



C. On Didijmium from different Sources. 

 By Professor C. M." Thompson, F.G.S. 



According to Kiesewetter and Kriiss (Berichte, xxi., 2,313), a solution of the 

 earths from Yttrotitanite from Arendal shows only three bands due to didymium. 

 On examining a moderately concentrated solution after separation from the bulk 

 of the other metals by precipitation with potassium sulphate, all the bands 

 shown by an ordinary didymium solution of similar strength were seen. 



Specimens of didymium from Gadolinite, from Orthite, and from Monazite were 

 also examined. No differences as compared with didymium from cerite could be 

 observed sufficiently marked to justify the inference that the bands varied in 

 strength in an independent manner. 



B B 2 



