Manchester Memoirs, Vol. Hi. (1907). 9 



continent, do not avoid an error inherent in the method 

 of heating the mixed gases together — viz., the gradual 

 combination of the gases as they are raised to the ignition 

 point. To obviate this difficulty, Falk has recently ignited 

 the mixture by sudden adiabatic compression in a steel 

 cylinder — and calculated the temperature from the volume 

 occupied at ignition. I think he could not avoid a local 

 cooling of the gas in contact with the cold walls of the 

 cylinder and piston, which would affect his calculated 

 ignition points. Mr. Coward has lately been working with 

 an apparatus I designed with the aid of Mr. G. W, A. 

 Foster, and has obtained ignition points which vary 

 through a comparatively small range of temperature. 

 The method consists in bringing the combustible gas 

 through a small tube passing along the axis of a large 

 tube, electrically heated. 



When, five years ago, Dr. Edgar and I began to 

 devise a method for burning a weighed quantity of 

 hydrogen in a weighed quantity of chlorine, we thought 

 our attempt — if it succeeded — would be regarded, at best, 

 as an academic vindication of Stas' atomic weight. We 

 did not anticipate that Richards in Harvard would be re- 

 determining the chlorine-silver ratio, and Guye in Geneva 

 and our countrymen, Scott and Gray, would be re- 

 determining nitrogen, and comparing it with silver and 

 chlorine. Silver was one of those few elements so exactly 

 determined that we wrote its atomic weight with six 

 figures, we knew it to a fraction of its ten-thousandth part ! 

 But the new determinations made it necessary either to 

 alter silver or chlorine : hence the sudden importance of 

 an independent determination of chlorine. Dr. Edgar has 

 repeated our experiments in a different manner, and 

 W. A. Noyes in America has made a fresh determination. 

 Guye has also redetermined the density of hydrogen 



