216 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



however, left much to be desired in the way of accuracy, and needed 

 verification. Therefore a very careful investigation, begun at Har- 

 vard with the assistance of Schumb, before the appearance of Fajan's 

 publication, furnished valuable knowledge on this point. 



Solubility of tiro kinds' of lead nitrate. 1 



Uranium 

 lead. 



Per cent salt in saturated solution (25.00°) 37.342 37. 2S0 



Grams lead per 100 grams water 37. 2S1 37. 130 



Molecular solubility per 1,000 grams water 1. 7993 1. 7989 



1 The uranium lead used in these determinations was a specimen from Australia having the atomic 

 weight 200.41, not quite like the earlier sample, but not different in important degree. 



Here, again, differences in weight alone are manifest, and these 

 are proportional to the differences in the atomic weights; the molec- 

 ular behavior is essentially identical in the two sorts. 



The identity in solubility might also be inferred from the impossi- 

 bility of separating the two kinds of lead from each other by frac- 

 tional crystallization. This was predicted by Soddy, and tested by 

 him and by others. Various vain attempts have been made to 

 separate the different kinds of lead from one another, but apparently 

 when once they are mixed, no ordinary method can separate them, 

 since the properties of the different kinds are so nearly alike. The 

 latest attempt at the Gibbs Memorial Laboratory involved 1,000 

 fractional crystallizations of the Australian lead nitrate, which is 

 believed to contain both ordinary and uranium-radium-lead. The 

 extreme fraction of the crystals (representing the least soluble por- 

 tion, if any difference in solubility might exist) gave within the 

 limit of error the same atomic weight as the extreme fraction of the 

 mother liquor (representing- the most soluble portion), thus con- 

 firming the work of others in this direction. 



When wires constructed of two different metals are joined, and the 

 junction heated, an electrical potential or electromotive force is pro- 

 duced at the junction. This property seemed, then, to be a highly 

 interesting one to test, in order to find out how great may be the 

 similarity of the two kinds of lead. In fact, wires made of radio- 

 active lead and ordinary lead tested in the Gibbs Laboratory gave 

 no measurable thermoelectric effect, the wires acting as if they were 

 made of the same identical substance, although the atomic weights 

 and densities were different. No other case of this sort is known, 

 so far as I am aware. The melting points of the two kinds of lead 

 were likewise found, with the assistance of N". F. Hall, to be identical 

 within the probable accuracy of the experiment, 1 



1 Since this address was first published Prof. P. W. Bridgman has found that both the 

 electrical conductivity of the two kinds of lead, and the effect of pressure on this con- 

 ductivity, are likewise identical ; and Prof. W. Duane has found that even the X-ray 

 spectra of the two kinds show no difference. 



