[mclennan-el worthy] jam in INTERFEROMETER 25 



Accuracy of the Determination 



The mean error of setting the cross hairs of the micrometer eye- 

 piece on the centre of a band was about 11-100 of the distance between 

 two bands. For mixtures up to 0-2% helium and 0-4% hydrogen 

 this would give an error of about 4% on the result. For percentages 

 up to 1% of either gas, the error due to this cause would be about 2% 

 on the result. Slight changes of position of the metal parts of the 

 interferometer, owing to temperature changes caused greater errors 

 on some occasions, manifested by a considerable zero shift during 

 readings, which could not always be allowed for. Such errors could be 

 made negligible by using longer interferometer tubes and by keeping 

 the instrument at a constant temperature. 



Altogether it was considered that determinations of percentages 

 in mixtures containing less than 1% hydrogen or helium were made 

 with a mean error of ±5%. 



Summary 



1. A method for the determination of small percentages of helium 

 or of hydrogen in air has been described, making use of a Jamin 

 interferometer. 



2. Two methods of calibrating the instrument have been out- 

 lined; one based on the relation between the difference in path, caused 

 by the displacement of air in one of the tubes by the mixture and the 

 resulting shift of bands; the second, a method based on the change of 

 refractive index of the air in one tube caused by lowering the pressure 

 on the air. 



3. With the instrument that was used the accuracy with which 

 determinations were made, was about ± 5% on the result. With 

 longer tubes a greater accuracy could be obtained. 



The work was carried out at the Admiralty Physical Laboratory, 

 South Kensington. 



April 15th, 1919. 



Sec. Ill, Sig. 3 



