320 . Scientific Intelligence. 



in which c, h and o represent the percentages of carbon, hydro- 

 gen and oxygen respectively. If a general expression cp = Ac 4- 

 Bh — Co be taken, the numerical value of the coefficient, A = 81, 

 must be maintained, since it corresponds to pure carbon, and all 

 known data (from 8060 to 8140) prove that the figure 81 must 

 really be taken for each per cent unit of carbon in the fuel (the 

 accuracy of the measurements being within the limits of from 1 

 to 2 per cent of the total heat of combustion.) For hydrogen, 

 however, the coefficient B == 345 cannot be maintained, because it 

 has been obtained from data relating to the burning of gaseous 

 hydrogen, while in ordinary liquid or solid fuel the elasticity of 

 the gas is lost. Its hydrogen must be considered as if liquefied 

 and consequently B must not exceed 300, supposing as usual that 

 the resulting water is in the liquid state. In order to find the 

 true coefficients suitable for practical purposes Me^deleeff 

 takes the value of cp = 4190, which is the correct value for cellu- 

 lose within one per cent, and is also the average of 79 most com- 

 plete determinations for fat coals (by Maler, Alexeyeff, Damski, 

 Diakonoff, Miklaschewski, Schwanhofer and Bunge)andthe aver- 

 age for naphtha fuel. From this he finds 



cp = 81c + 300A — 26(o-s) 



(s being sulphur). This formula represents with an accuracy of 

 trom 1 to 2 per cent the heat of combustion of pure charcoal, 

 coke, coals, lignites, wood, cellulose and naphtha fuels. It applies 

 of course to the best determinations only ; especially to those which 

 have been made in a calori metric bomb where the error is less 

 than 1 or 2 per cent. This formula of course is only an approxi- 

 mate empirical expression of facts; but it corresponds at the same 

 time to the numerical value of the coefficient B for hydrogen 

 which theoretical considerations would lead us to expect. — J. 

 Chem. and Phys. Soc. Husse, xxix, 144; Nature, lvi, 186, June r 

 1897. G. F. B. 



5. The capillary constants of molten metals. — This is the sub- 

 ject of an inaugural dissertation (Gottingen, 1897) by Henry 

 Siedentopf. The author has employed Quincke's method of ob- 

 taining the capillary constants by means of falling drops (1868). 

 Whereas Quincke, however, depended upon the determination of 

 the weight of the drops, the present author has based his results 

 upon the measurement of the curvature of their surface and their 

 size. The metals experimented upon were cadmium, tin, lead, 

 mercury and bismuth. For each the surface tension and specific 

 cohesion were determined at a temperature near that of fusion. 

 The results are given in the following table : 



Temperature Surface Specific 



of fusion. tension. cohesion. 



Cadmium.. 318° 84*85 2125 



Tin _ 226° 62-43 17*87 



Lead 325° 51-94 9-778 



Mercury. —39° 46*29 6'767 



Bismuth 264° 43*78 8'755 



