178 RADIATION BIOLOGY 



to that produced in the gas (air) of the thimble chamber varies with the 

 atomic number of the chamber material, its volume, and the energy of the 

 photons. It has already been shown that this effect is small and calculable 

 under certain conditions, but it must be anticipated that no air-equivalent 

 wall remains truly so anywhere from 1 to 3000 kev. The regions of pho- 

 ton energy most difficult to cover are those in which both the Compton 

 and photoelectric process contribute substantially to the release of elec- 

 trons from the wall, namely, the range below 50 kev.^^ At very low volt- 

 ages absorption in the chamber wall is the predominant cause of discrep- 

 ancy. By cut-and-try method, however, the material, shape, and posi- 

 tion of the walls and collector electrode may be adjusted so that the 

 electrons released produce the same net effect as air. It is thus possible 

 to make a chamber having a graphite shell with aluminum collector which 

 will give readings nearly proportional to the standard chamber over a 

 wide range of the harder photon energies. Without going into further 

 detail, it is evident that a thimble (air-wall) chamber cannot be relied 

 upon to indicate true roentgens unless calibrated against a standard open- 

 air chamber. This is usually done with an X-ray tube of small focal spot, 

 operated under reproducible and steady conditions of voltage and current. 

 The dosage rate of the X-ray beam, after suitable collimation by dia- 

 phragms, is measured in roentgens per minute at a given point with a 

 standard ionization chamber. Holding the operation of the X-ray tube 

 steady, the standard chamber is removed from the beam, the center of the 

 thimble chamber is placed at the exact position of the front edge of the 

 limiting diaphragm (D, Fig. 2-2) of the standard chamber, and the time- 

 rate response is noted for the thimble-chamber indicator. This must be 

 done for each quality of radiation for which the instrument is to be used. 

 When the dependence of the response upon quality is small, only a few 

 calibrations are necessary, it then being permissible to interpolate for 

 intermediate qualities. Experience has shown that it is seldom safe to 

 extrapolate such calibration. 



Several other sources of error must be guarded against (X-ray Stand- 

 ards and Units, 1934) ; for example, there may be some uncertainty as to 

 the effective center of a thimble chamber when it is placed in the test 

 beam. To avoid undue errors from this source a thimble chamber should 

 not be calibrated nearer to the tube target than a distance 50 to 100 times 

 its radius. It must likewise be assured by test that the X-ray beam is of 

 uniform intensity over an area of diameter somewhat larger than both the 

 linear dimensions of the thimble chamber and the diameter of the dia- 

 phragm on the standard. Moreover, filters necessary in the calibration 

 should be so placed that the effect of secondary radiation therefrom is not 

 included in the measurements. Whenever the inverse square law is used 



13 The term "kev" will be reserved to denote the energy of the photon, whereas the 

 term "kv" will refer to the voltage applied to an X-ray tube. 



