Mr. R. Mallet on Expansion by Refrigeration. 233 



portion was much less dense than the exterior, the opposite of what 

 must have occurred had expansion in volume on cooling taken place. 



It is a fact, notwithstanding what precedes, and is well known 

 to ironfounders, that certain pieces of cold cast iron do float on 

 molten cast iron of the same quality, though they cannot do so 

 through their buoyancy. As various sorts of cast iron vary in 

 specific gravity at 60° F., from nearly 7*700 down to 6-300, and 

 vary also in dilatability, some cast irons may thus float or sink in 

 molten cast iron of different qualities from themselves through 

 buoyancy or negative buoyancy alone ; but where the cold cast iron 

 floats upon molten cast iron of less specific gravity than itself, the 

 author shows that some other force, the nature of which yet re- 

 mains to be investigated, keeps it floating ; this the author has pro- 

 visionally called the repellent force, and has shown that its amount 

 is, cceteris paribus, dependent upon the relation that subsists be- 

 tween the volume and "effective" surface of the floating piece. 

 By " effective " surface is meant all such part of the immersed solid 

 as is in a horizontal plane or can be reduced to one. The repellent 

 force has also relations to the difference in temperature between 

 the solid and the molten metal on which it floats. 



The author then extends his experiments to ]ead, a metal known 

 to contract greatly in solidifying, and, with respect to which, no one 

 has suggested that it expands at the moment of consolidation. He 

 finds that pieces of lead having a specific gravity of 11-361, and 

 being at 70° F., float or sink upon molten lead of the same quality, 

 whose calculated specific gravity was 11*07, according to the rela- 

 tion that subsists between the volume and the " effective " surface 

 of the solid piece, thin pieces with large surface always floating, 

 and vice versa. An explanation is offered of the true cause of the 

 ascending and descending currents observed in very large "ladles" 

 of liquid cast iron, as stated by Messrs. Nasmyth and Carpenter. 

 The facts are shown to be in accordance with those above men- 

 tioned, and when rightly interpreted to be at variance with the 

 views of these authors. 



Lastly, the author proceeds to examine the statements made by 

 these writers, as to the floating of lumps of solidified iron furnace- 

 slag upon the same when in a molten state • he examines the condi- 

 tions of the alleged facts, and refers to his own experiments upon 

 the total contraction of such slags, made at Barrow Iron-works (a 

 full account of which he has given in his paper on "The true 

 Nature and Origin of Volcanic Heat and Energy," printed in Phil. 

 Trans. 1873), as conclusively proving that such slags are not denser 

 in the molten than in the solid state, and that the floating referred 

 to is due to other causes. The author returns thanks to several 

 persons for facilities liberally afforded him in making these expe- 

 riments. 



"Spectroscopic Notes. — No. I. On the Absorption of great 

 Thicknesses of Metallic and Metalloidal "Vapours." By J. Norman 

 Lockyer, F.E.S. 



It has been assumed hitherto that a great thickness of a gas or 



