646 



SCIENCE 



[N. S. Vol. XLVI. No. 1200 



operated by a six-volt storage cell. It may be 

 lowered into a hollow object, the lamp being at- 

 tached to the microscope tube and moving with 

 it. Especially convenient for the study of enam- 

 els, alloys, opaque objects and substances con- 

 tained in opaque vessels. A model will be ex- 

 hibited in operation. 



The qualitative separation and detection of gal- 

 lium: Philip E. Browning and Lyman E. 

 Porter. A study of the occurrence of the element 

 shows it to be most closely associated with Pb, Al, 

 Fe, Mn, Zn and In. Analytically it falls into the 

 Al group, its hydroxide being precipitated by 

 NHjOH in the presence of NH4CI and being sol- 

 uble in an excess of NaOH. The chief analytical 

 problem is its separation from Al and two methods 

 are studied, both of which give satisfactory re- 

 sults. First, the method of de Bois Vaudran, pre- 

 cipitating Ge3(FeC6No)i by KiFeCoNo in the pres- 

 ence of strong HCl to about one third the volume 

 of the liquid. Second, saturating a solution with 

 HClga in the presence of ether, which throws out 

 the AICI3 and keeps the Ga in solution. 



The qualitative detection of germanium and its 

 separation from arsenic: Philip E. Browning 

 and Sewell E. Scott. A study of the occurrence 

 of the element shows it to be most closely associ- 

 ated with Ag, Pb, Hg, Cd, As, Sn, Zn, Ti and Cb. 

 It falls in the analytical group with As and Sn 

 since its sulphide is soluble in (NH4)2S. It is 

 separated from Sn by treating the sulphides with 

 (NH.,)2C03, GeSj being soluble. Prom As it may 

 be separated by treating a solution of the sulpho 

 salts with ammonium acetate, acidifying with 

 acetic acid and passing H,S. AS2S3 is precipi- 

 tated and Ge remains in solution. The following 

 modification of Buchanan's method was devised 

 for the separation and detection of Ge. The 

 germanium material was dissolved in strong hydro- 

 chloric acid (5-10 cm.^) in a test tube some 

 KMnOj added, to keep arsenic if present in the 

 higher condition of oxidation and distilled into 

 another test tube kept cool in water. After dis- 

 tilling about one half volume the Ge is found in 

 the distillate by means of H.S. 



Silver anion : H. C. P. Weber. It is customary to 

 think of silver as a strictly monovalent element, 

 which forms in solution a positive ion. When a 

 solution of a silver salt is electrolyzed at high 

 current density a black deposit is formed at the 

 anode which has been variously described as silver 

 peroxide and as silver perosyuitrate, the formulas 

 ascribed varying but tending to indicate the pres- 



ence of trivalent silver. It is now shovm that in 

 this compound we have silver which in transference 

 experiments acts as an anion, probably trivalent, 

 a very unstable and intensely active oxidizing 

 agent. It is not derived from hydrogen peroxide 

 but rather of the permanganate type. The com- 

 pound is of great interest in connection with the 

 valence of silver in particular, and valence in 

 general. 



The fixation of nitrogen with the silent electric 

 discharge: Farrington Daniels and Oliver E. 

 WuLF. The oxidation of nitrogen by the silent or 

 cold electric discharge has been proved. No 

 energy is lost as heat, and under the proper con- 

 ditions nitrogen pentoxide instead of nitrogen 

 peroxide is formed. This should simplify the ab- 

 sorption towers. Pressure favors this reaction 

 but not the reaction which gives nitric oxide. 

 Practical applications have failed because the re- 

 action is too slow. A search for a catalyzer was 

 unsuccessful. Experiments with various types of 

 discharge chambers look hopeful. 



The displacement of nitric hy carbonic acid in 

 silver nitrate solutions and the relation of this re- 

 action to the inclusion error in the silver voltam- 

 eter : A. S. McDaniel and H. D. Hineline. It 

 has been shown that carbon dioxide reacts slowly 

 with silver nitrate in aqueous solution forming a 

 carbonate of silver and liberating free nitric acid. 

 Ci'ystals of the carbonate have been isolated and 

 identified. The nitric acid liberated has been 

 estimated by titration with iod-eosin and its 

 amount compared with the silver contained in the 

 crystals of silver carbonate. The reaction is be- 

 lieved to be as follows: 



Ag NO3 + H2CO3 ?± Ag HCO3 -f HNO3. 



About one one-hundredth of one per cent, of the 

 silver nitrate is converted to the carbonate. In 

 the silver voltameter a clear solution of silver ni- 

 trate which has been saturated with CO2 gives a 

 deposit about 0.4 per cent, too heavy. This effect 

 was first shown by Rosa Vinal and McDaniel, but 

 it was thought by them that the amount of COj 

 normally present in air has no appreciable effect 

 upon the mass of deposit. In the present investi- 

 gation a few direct measurements have been made 

 of the effects produced by one to ten times the 

 normal amounts of CO, present in the air and 

 while the results are incomplete they indicate that 

 the effect of the normal amount of CO2 in the air 

 is not negligible and indeed may be larger than 

 the inclusion error in normal deposits. 

 {To te continued) 



