366 On the Fluid Density of certain Metals. 



be best understood by a short description of the way in which 

 the metal tin was treated by them. 



Tin was melted in a suitable vessel, the melting-point (viz. 

 226°'5 C.) being carefully maintained. Pieces of tin were 

 then dropped in, and observed first to sink, then to rise before 

 melting — showing that solid tin has a higher specific gravity 

 than the liquid mass, but that when its temperature has been 

 raised to the melting-point, or rather to a certain temperature 

 just below, then it is specifically lighter. In order to deter- 

 mine this relation, they enclosed pieces of copper (having a 

 higher specific gravity than tin) in pieces of tin, and were 

 thus able to increase the specific gravity of the solid piece at 

 will; and although not able to determine directly whether the 

 compound piece was of the same density as the liquid, the 

 latter not being transparent, yet by varying the amount of 

 copper enclosed in the tin, thus forming pieces of a different 

 density, they were enabled to say that it lay between the 

 two limiting values. Compound pieces with varying and 

 known amounts of copper, thus formed, were placed on the 

 surface of the liquid tin for some little time, then dipped 

 under the surface; and it was noted whether they rose again 

 or not. The pieces were each then carefully taken out again 

 and weighed, to ascertain that no part had been melted. 

 This was continued until they found two compound pieces 

 such that, at a temperature just below the melting-point, one 

 would just rise and the other would just sink. 



If, then, the relative densities of the liquid tin and the copper 

 at the temperature of liquid tin could be ascertained by expe- 

 riment, the relative density of the solid tin just below melting- 

 point and the liquid tin can be calculated. 



A copper ball was heated to the temperature of the melting 

 tin, and weighed, first in air, then in the liquid tin. The rela- 

 tive densities being thus ascertained, the calculations, which 

 we need not give here, are of a simpler character. 



In the case of tin, they found that the liquid is of *7 per cent, 

 greater density than the hot solid at a temperature just below 

 melting-point. 



This, so far as it goes, is confirmatory of the results obtained 

 by us. If diagram No 2. be examined, it will be seen that the 

 line of volume rises gradually to the line of equilibrium, indi- 

 dicating expansion; and just before the ball melts rises above 

 the line, indicating a floating effect, when the temperature 

 approximates to the melting-point. 



Tin therefore appears to be similar to iron ; viz. it is at 

 its maximum density when cold, and at its minimum density 

 when at a temperature just below melting, and that the fluid- 

 density is between the two. 



