660 PROCEEDINGS OP THE AMERICAN ACADEMY. 



(2) Stir with plenty of Hg, repeat till fresh Hg is not tarnished. 

 Decant. 



(3) Stir with strong aqueous solution of HgCl 2 till no more precip- 

 itate, remove HgCl 2 (separating funnel), keep in dark. 



(4) Dry over fused CaCl 2 at least a day. Stir. Keep cool and dark. 



(5) Decant, add P 2 5 (resublimed if necessary to remove P), digest 

 for some hours (with return condenser), distil. 



In the last step the distillate before condensation was passed through 

 about four feet (120 cm.) of drying tubes filled with phosphoric pen- 

 toxide and glass wool, and kept at about 50° C. It was then washed 

 free from any possible particles of pentoxide by bubbling up through 

 some ten inches (25 cm.) of chemically clean glass beads immersed in 

 a portion of the liquid earlier distilled. The pure dry gas was finally 

 condensed and run into the receiving bottle. The bottle had already 

 been connected to the observing tube so that the liquid could be used 

 without exposure to the undried and more or less dusty air of the room. 

 All joints were of ground glass mercury-sealed. The temperature of 

 the vapor during distillation did not vary more than 0°.02 till near the 

 close, when the vapor began to superheat on the warmer walls of the 

 distilling bottle. Carbon bisulphide thus purified is a beautiful, spar- 

 kling, colorless, limpid liquid with a pleasant odor suggesting ether. 

 Probably the most delicate test of its purity is its odor. The bad smell 

 of commercial carbon bisulphide is caused by the impurities. 



In the purification of this substance in such large quantities as this, — 

 about twelve liters at a time, — every precaution was taken against fire. 

 No flames were used in the room, and the liquid was distilled from a 

 water-bath warmed by an immersed electric heater. Thus there was no 

 object as hot as boiling water, and no spark, which could possibly be 

 reached by any vapor of carbon bisulphide which might escape. 



Such are the method and devices developed step by step by actual 

 experience in solving this threefold problem, — optical, electrical, and 

 chemical. 



Procedure. 



In these measurements the carbon bisulphide was purified as de- 

 scribed, but had stood in glass bottles for several months, though in the 

 dark, hermetically sealed, and dry. After the measurements it was 

 examined, and its freedom from disagreeable odor made it seem probable 

 that deterioration had not taken place. 



Each observation consisted essentially of three photographs of the 



