Fluid Density of certain Metals. 



365 



the five metals in which both systems were tried, the results 

 correspond as closely as could be expected, taking into con- 

 sideration the great difficulties of observation at the high 

 temperatures employed. We therefore venture to think that 

 the oncosimeter can be depended upon generally for the deter- 

 mination of the fluid densities of metals, if the experiments are 

 conducted with the necessaiy care. 



Taking therefore the oncosimeter-results as approximately 

 accurate, we find the change of volume of these metals in pass- 

 ing from the cold solid to the liquid state to be as follows : — 



Metal. 



Specific gravity ] Specific gravity 



of solid. 



of liquid. 



Bismuth 

 Copper 

 Lead . . . 



Tin 



Zinc. 



Silver ... 



*Iron (No. 4, foundry, "1 

 Cleveland) f 



9-82 



8-8 

 11-4 



7-5 



72 

 10-57 



6-95 



10-055 

 8-217 



10-37 

 7-025 

 6-48 

 9-51 



Percentage of change 



in volume from cold 



solid to liquid. 



Decrease of vol. 23 

 Increase of vol. 7*1 

 do. do. 9-93 



do. 



do. 



do. 



do. 

 do. 

 do. 



6-76 

 11-1 

 11-2 



do. do. 



1-02 



The experiments just described lead us to believe that, 

 although the conditions under which they are made may 

 render it impossible to obtain results that are rigidly accurate, 

 still the errors are inconsiderable. It has been urged that the 

 unsoundness or porosity in the casting would disturb the accu- 

 racy of the results ; but this is provided against in each case 

 by taking the specific gravity of the ball operated on, its iron 

 stem being submerged in water to a point which is afterwards 

 just reached by the molten metal. It has also been suggested 

 that the expansion of the ball, when it enters the fluid metal, 

 causes tension on the metal in the interior of the ball, and that 

 therefore the density of the ball (from which the fluid density 

 is determined) can never be trusted. It must be remembered, 

 however, that the determination of the fluid density is made 

 and automatically registered at the moment of immersion, 

 before the volume of the ball can be practically affected. 



nice our 



last 



paper 



read, 



have 



that MM. F. 



Nies and A.Winkelmannt have been investigating the changes 

 in volume of certain metals when melted. Their method will 



* Wrightson, Journal of the Iron and Steel Institute, No. 1 (1880), p. 20. 

 t SitzungsbericMe der Akademie der Wissenschaften zu Milnchen, 1881, 

 part 1, p. 03. 



