586 FRED M. UBER 



spectrometer {1, p. 38). Since it is practically impossible to avoid 

 "having some air in the instrument, such corrections are usually indi- 

 cated. When N^^ is being measured, air is shown to be present by 

 the occurrence of an oxygen peak at mass 32 and confirmed if an argon 

 peak appears at mass 40. Although the ratio of nitrogen to oxygen 

 in the atmosphere is approximately four, the air nitrogen contribu- 

 tion to the mass 28 peak by a contaminated test sample is usually 

 greater than would be indicated by multiplying the height of the 

 mass 32 peak by four. The ratio as actually measured depends on 

 instrumental and other factors, including differential ionization prob- 

 abilities for the two elements. A proper correction factor, charac- 

 teristic for each mass spectrometer, must be determined so that cor- 

 rections can be applied to the readings at the various mass numbers, 

 should air leaks be encountered. 



Instead of expressing isotope concentrations in atom per cent as 

 calculated by the foregoing equations, there is an increasing tendency 

 to publish results in terms of the atom per cent excess concentration as 

 compared to the normal abundance or to a laboratory standard. 

 These excess values are computed by subtracting either the normal 

 concentration, as tabulated in Table I, or the concentration of an ar- 

 bitrary, but permanent, laboratory standard. The latter procedure 

 is preferable for several reasons, including a continual check on the 

 performance of the mass spectrometer and on any variations in the 

 concentration of the normal isotope. Some types of calculations, 

 as used for example with the isotope dilution technique, are simplified 

 when concentrations are expressed in terms of atom per cent excess. 

 Since arbitrary standards of some elements prepared from chemical 

 salts may possess higher concentrations than occur normally in bio- 

 logical preparations, negative values will be encountered occasionally 



a, p. 28). 



D. LIMITATIONS AND SOURCES OF ERROR 

 . 1. Absolute Accuracy of Isotope Concentrations 



To achieve high absolute accuracy in stable isotope measurements 

 with a mass spectrometer is very difficult. The sources of the errors 

 are inherent in the design of the instrument itself and cannot be elim- 

 inated easily. Fortunately most of the biological applications of 

 stable isotopes are based on a determination of their relative abun- 



