2o8 REPORTS O^ INVESTIGATIONS AND PROJECTS. 



(CeH,COO)C(CeH,ONa),< > 



Colorless. 



NaOOCCgH.CC : CgH, : 0)(C,H,ONa)< > 



Colored. 



NaOOCCeH,C( : C^H, : O) (CgH^ONa) 

 Deeply colored. 



An Electrically-Controlled Gas Regulator. By E. E. Reid. (American 

 Chemical Journal, 41, p. 148.) — This article describes a very efficient iron gas 

 regulator, which is cheap and simple in construction. 



An Apparatus for the Purification of Mercury. By L. Junius Desha. 

 (American Chemical Journal, 41, p. 152.) — Mercury is automatically forced 

 in fine streams through dilute nitric acid about 150 times a day, and is elec- 

 trolyzed at the same time. This procedure removes with ease practically all 

 of the metallic impurities from the mercury, which is then distilled. The 

 apparatus is very simple and runs day and night without attention. 



Bancroft, Wilder D., Cornell University, Ithaca, New York. Grant No. 

 556. Systematic study of alloys. (For previous reports see Year Books 

 Nos. 2, 3, 4, 5, 6, and 7.) $i,500- 



The iron-carbon alloys have been annealed in a vacuum until equilibrium 

 was reached, and have then been examined under the microscope. The re- 

 sults show that we are dealing with a reversible equilibrium and that the 

 Upton diagram is substantially correct. We are now running heating and 

 cooling curves in order to determine the different inversion temperatures. 

 We hope also to show the existence of 8 iron and thus to confirm the work 

 of Ball, of Osmond, and of Pierre Curie. 



Tensile-strength tests have been made on the zinc-aluminum alloys and we 

 are now investigating the cause of the disintegration of some of these alloys. 



The diagram for the zinc-antimony alloys, as published by Monkemeyer 

 some years ago, contained a line which was theoretically impossible. This 

 difficulty has now been cleared up. It was due in large part to hysteresis 

 phenomena during cooling. Errors of 70° or even 100° in the determination 

 of inversion temperatures may occur when cooling curves alone are taken. 

 This case illustrates in a striking manner the possibility of error due to super- 

 cooling. We have also determined the diagrams for several other binary 

 alloys containing zinc as one component. A preliminary account of this work 

 will appear in the Journal of Physical Chemistry for November. 



