X-RAY MICROSCOPY 



Fig. 18. Photomicrographed from a contact 

 microradiogram of human erythrocyte. 



smeared sample is shown in Fig. 18. The 

 microradiogram, after exposure, was washed 

 in acetone to remove the sample and Parlo- 

 dion film and was developed for three min- 

 utes in D-158. 0-K(23.7 A) radiation was 

 used as obtained from an oxidized copper 

 anode target excited at one kilovolt and with 

 a chromiimi filter of transmission as shown 

 in Fig. 8. 



This ultrasoft radiation yields a value for 

 7/im of about 1.2, and with 7 equal to 1.72, 

 juw is about 0.7. From (15) and for (y/N) 

 equal to about 60 for this measurement, the 

 predicted average relative error is 3% in a 

 mass per unit area, m, determination. 



The total cell mass was determined as il- 

 lustrated in Fig. 19. Radial symmetry was 

 assumed in these measurements and sixteen 

 photometer readings, p2 , and the associated 

 background reading, pi , were taken from an 

 averaged profile as determined, from the mi- 



AVERAGED PROFILE 



ERYTHROCYTE 



DATA RCOUCHON GRID 



1/2 » I /I SLIT MEASUREMENT 



Fig. 19. Photometer readings, p2 and pi , are 

 taken from an average profile, assuming radial 

 symmetry, at sixteen points at radial positions of 

 such interval as to yield equal effect upon the total 

 mass measurement. The reduction of such data to 

 mass per unit area, m, and subsequent numerical 

 integration for total mass was carried out with a 

 digital computer. 



.75 1.00 



Fig. 20. The calculated values of yum for eight 

 hundred single cell measurements were divided by 

 the average value to obtain a corresponding num- 

 ber of m/fFi values which are thus independent of 

 7 and M and their respective errors . These data were 

 counted in intervals of 0.1. The precision (average 

 deviation) of a single measurement is of the order 

 0.03. 



692 



