230 



RADIATION HIOI.OGY 



ol)viously should ncxcr he incjisuiccl as a whole ohjeci. The ei'ior in 

 moasuriii^ spurscly clistrihutcd chiomosomcs on a mctaphasc plate 

 would j^orhaps he comparahle with this. As the heterogeneity decreases, 

 and the darker and lighter parts of the object come to contain more nearly 

 Hke amounts of absorbing maforial, the distributional error rapidly 

 decreases. For an extensive theoretical treatment of the distributional 

 error see Ornstein (1952) who considers several methods of minimizing or 

 correcting for the error. 



There is an error involved in the summation of transmissions where 

 intensity distribution is uneven, and also an error in converting the total 

 or average transmission of a heterogeneous object into an average extinc- 

 tion. It appears that, in direction and relative magnitude, these should 



run parallel w'ith the distributional 

 eri'or which w^as just discussed (see 

 Fano, 1947; Swift, 1950). 



The unexceptionable method for de- 

 termining amount of absorbing mate- 

 rial in large areas with varying density 

 is that of making densitometer cross- 

 tracings of negatives (Caspersson, 

 1940a) or a series of scannings of a cell, 

 in many azimuths, with the photo- 

 metric apparatus (Thorell, 1947). 

 Caspersson followed this procedure 

 with living early meiotic prophase 

 nuclei of the grasshopper; the cross- 

 trace in one azimuth is shown in Fig. 

 6-1 1 . He did not detect any distortion 

 of what was to be expected from absorp- 

 tion of a sphere aside from some margi- 

 nal diffraction, i.e., there was no meas- 

 urable heterogeneity although, visibly, 

 there seemed somedensei' structures. Accordingly, he used the extinction 

 through the center as a measure of concentration, and computed total 

 amount in the nucleus from this datum. 



Other possible sources of error in microspectrophotometry of cells are 

 discussed in such references as Caspersson (193(), 1940a, b, 1950) ; Thorell 

 (1947); Swift (1950, 1953); Naora (1951); PoUister (1952a); and Davies 

 and Walker (1953). 



20 30 



MICRONS 



40 



50 



Fig. 6-11. intraviolot absorption 

 measureiucnts at a series of points 

 across the diameter of a grass- 

 hopper spermatocyte (soHd line), 

 compared with the computed ab- 

 sorption (uirve of an absorbing 

 sphere (broken line). (Redranm 

 from Caspersson, 1939.) 



4-4. QUANTITATIVE APPLIC.\TIONS, ABSOLUTE AND REL.\TIVE 



Because of the potential errors in photometric analysis of cells it is 

 evident that a straightforward computation of absolute amount of sub- 

 stance by referring light loss in a cell to any standard value obtained on 



