GENERAL PROCEDURES FOR RADIOASSAY 181 



With this information it becomes possible to choose between different 

 methods for estimation and application of the self-absorption correction. 



Thin Samples. The self-absorption loss may be insignificant if the 

 sample mass is sufficiently low. As may be noted from Fig. 5-8, the self- 

 absorption loss from Ca*^ amounts to about 5 per cent at a sample thick- 

 ness of 1 mg/cm-. This means that, depending on the accuracy recjuired, 

 the Ca^^ samples must not exceed about O.o mg/cm- when it is desired to 

 neglect the self-absorption loss. This method can therefore be employed 

 only when the specific activity of the sample is high enough and the meas- 

 urement sufficiently sensitive so that a reasonable count can be obtained 

 with the limited mass of the sample. At thicknesses of less than 0.5 mg/ 

 cm", there are uncertainties in the curve due to self-scattering phenomena. 

 However, these errors are within the usual biological uncertainties, so that 

 the limiting factor in the use of this procedure is the specific activity 

 obtainable in the sample to be measured. 



Samples of Standard Thickness. Self-absorption corrections can be 

 eliminated if it is possible to prepare all the samples and standards of 

 identical thickness and area. This may often be accomplished for rou- 

 tine methods by using standardized amounts of carrier that are large com- 

 pared with the amount of the element in the original samples. With 

 calcium in animal samples, for example, this would be satisfactory for the 

 soft tissues but not generally for skeletal tissues or feces, which themselves 

 contain large amounts of this element. This procedure could also be 

 accomplished by preliminary chemical analysis, the results therefrom 

 being used to choose a specific sample alic^uot to yield the standard 

 thickness. 



Empirical Relationships. The most general and widely used procedure 

 is based on the empirical relationship between the thickness of the sample 

 and the degree of self-absorption under the conditions of measurement. 

 This relationship is expressed for Ca^^ in Fig. 5-8. Thus, for instance, a 

 sample of 10 mg/cm- would register about 58 per cent of its counts under 

 an end-window counter as compared with a standard of zero mass. A 

 table of correction factors can be prepared from the curve, and each sam- 

 ple can be corrected on the basis of its mass. Although it is recommended 

 that each individual investigator prepare his own self-absorption curves, 

 Comar et al. (32) pointed out that with Ca^* such curves for end-window 

 counters were similar as determined in different laboratories. Further- 

 more the curves were relatively unaffected by the shelf position, the com- 

 position of the sample, or the window thickness between 1.8 and 3.5 mg/ 

 cm'-. This is of practical importance in the laboratory, since it means 

 that some of the unavoidable variations in procedure (e.g., replacing a 

 worn-out counter tube) will not necessitate the establishment of a new 

 self-absorption curve. However, as shown in Fig. 5-8, larger corrections 



