2l6 



SCIENCE PROGRESS 



necessary to evaluate it, as there are so many lines whose masses 

 are known which can be therefore taken as standards. What 

 is actually done is to take a fixed point called the register spot, 

 which is photographed as a circular dot on each spectrum (see 

 Fig. 2), and measure the relative distances of known and 

 unknown lines from this. By altering the magnetic field, 

 a chosen group of known lines can be photographed in various 

 positions. In this way the plate can be fully calibrated, a 

 process which is assisted materially by the fact that at the 



VII 



Kttn 



TW 



n n n 

 1.1 I 



dJU 



ct 



H H n 

 t 1 1 



III 



dL 



n n n 



. i I 



/ II 



viiin 



S9 



._ CO 00 00 



i-^ 





I / 



fill 



iKr 



_i_ 



I I I 



3xe 



IX 



I 



ffi 



/// 



Fig. 2. — Typical mass-spectra. 



Reproduced by kind permission of the Philosophical Magazine. 



end corresponding to greatest deflection — the left-hand end in 

 the illustrations, — ^the relation between position and mass 

 happens to be nearly linear. 



An excellent gas for calibration is a mixture of carbon 

 dioxide and methane. Taking the brighter lines given by this 

 mixture in order (omitting the hydrogen atom and molecule), 

 we have first a very well marked and easily recognised group 

 called the Ci group : this consists of 12 — C, 13 — CH, 14 — CH,, 

 15 — CHs, 16 — CH4 or O. After a gap this is followed by 

 the C2 group 24, 25, 26, 27, 28, 29, 30, starting with what 



