August 17, 1906.] 



SCIENCE. 



205 



tral solutions behaves qualitatively exactly 

 like iron oxidized in the flame. Its ca- 

 pacity, however, is somewhat larger than 

 that of the passive iron, indicating a thin 

 film or sheet. The resistance of this sheet 

 is relatively smaller, due, as we suppose, 

 to small holes or crevasses in the sheet. 



The Electrical Conductivity of Tungstate 

 Solutio'iis: Roger Clark Wells. 

 A study of the conductivity of various 

 tungstates showed that in the case of so- 

 dium metatungstate and ammonium para- 

 tungstate a partial transposition begins to 

 take place as soon as those salts are dis- 

 solved in water. Although the rate of this 

 transformation is very slow at 25°, it in- 

 creases rapidly with rising temperature. 

 This discovery will undoubtedly explain 

 the queer solubility determinations which 

 several investigators have found for these 

 salts without considering the time factor. 



Freezing-point Measurements: W. G. Smea- 



TON. 



The author takes up a consideration of 

 the difficulties encountered in freezing- 

 point determinations and then proposes a 

 method which is a modification of the Ra- 

 oult method. This modification is based 

 on the fact that, although cryohydrates are 

 theoretically ideal cooling baths, in prac- 

 tise their use necessitates the introduction 

 of the Nemst and Newton constants. The 

 use of these constants is made necessary 

 because the temperature of the cooling bath 

 can not be regulated to produce a tempera- 

 ture equilibrium in the freezing-point ves- 

 sel at the apparent freezing-point of the 

 solution in all cases. In applying the 

 modification it is most convenient to use a 

 mixture of ice and salt. An auxiliary cool- 

 ing bath permits rapid determinations. 

 The apparent freezing-point is first de- 

 termined rapidly in the auxiliary cooling 

 bath. Then the temperature of the cooling 

 bath is regulated to give temperature equi- 



librium at the apparent freezing-point. 

 In the meantime the ice has been thawed 

 out of the freezing-point vessel which is 

 undercooled in the auxiliary bath and then 

 is transferred to the other bath. Inocula- 

 tion is made when the temperature begins 

 to rise uniformly. Thus the degree of 

 undercooling is determined with the maxi- 

 mum of accuracy. Ice formation is pre- 

 vented during undercooling by vigorous 

 hand stirring. The method is rapid, easily 

 manipulated, and gives accurate determina- 

 tions with very small volumes of solution. 

 The only correction to be applied arises 

 from the change in concentration of the 

 solution through the ice separated. The 

 factor to be applied is a constant for a 

 given apparatus under uniform conditions. 



On Amorphous Sulphur; IV. Precipitated 

 Sulphur: Alexander Smith and R. H. 

 Brownlee. 



This investigation deals with the propor- 

 tions of amorphous sulphur ('supercooled 

 S/i') contained in sulphur which has been 

 precipitated (1) from sodium polysulphide 

 by the action of acids and of iodine, and 

 (2) from sodium thiosulphate by the ac- 

 tion of equivalent and excessive amounts 

 of acids. The sulphur from polysulphides 

 —so-called 'amorphous sulphur'— is almost 

 wholly crystalline soluble sulphur. When 

 the thiosulphate is used, different acids in 

 equivalent concentrations give different 

 proportions of amorphous sulphur. For a 

 single acid the proportion increases more 

 rapidly than the concentration of the acid. 

 The proportion of insoluble sulphur seems 

 to be greater the more rapid the action 

 (due to high concentration of the acid), 

 and therefore the smaller the droplets and 

 the quicker the hardening of the precipi- 

 tated liquid S/x,. Higher temperatures up 

 to 25° hasten the action, and therefore give 

 larger proportions of amorphous sulphur; 

 but at 40° the tendency of S/x, to revert to 



