A CRITIQUE OF CYTOCHEMICAL METHODS 



229 



tion of microspectrophotometry considerably more difficult than in the 

 simplification suggested. He believes that if the intensity distribution 

 in the microscopic image is to correspond in every detail with that in the 

 cell, a requirement for absorption measurements, then the demands on 

 the microscope are essentiall}^ identi('al with those set forth by Abbe for 

 highest resolution. Others have suggested that an absorption micro- 

 scope can perhaps be a compromise between the simplified optical system 

 of such instruments as colorimeters and that for the sharpest images at 

 high magnification (Norris et al., 1951; Grey, 1952; Kavanagh, 1952). 



4-3. SOME ERRORS OF QUANTITATIVE 

 MICROSPECTROPHOTOMETRY 



In practice, an apparatus for microscopic absorption is indeed much like 

 a colorimeter or a spectrophotometer, except for the introduction of a 

 microscope into the optical pathway, and the actual absorption measure- 

 ment is essentially the same — Ix is 

 intensity measured through the cell 

 while /o is a second reading through 

 an empty part of the slide, outside the 

 section. Such quantitative absorp- 

 tion data are easily obtained, but the 

 successful evaluation of the results 

 must take into consideration many 

 possible sources of error which arise 

 from the nature of cytological mate- 

 rial and the fact that the microscope 

 is used. Most substances within the 

 cell are in a physical state very differ- 

 ent from the dilute solutions meas- 

 ured in a colorimeter or spectropho- 

 tometer. The proteins and nucleic 

 acids are very concentrated and if 

 fixed, possibly even when unfixed, are 

 more like solid precipitates or gels 

 than solutions. Very little is actu- 

 ally known about the extent to which 

 this physical state can affect the 

 operation of Beer's law because no 

 extracellular model for such a study 

 is available. Solutions of nucleic acid which approach that which occurs 

 in cells (1-5 per cent) appear to give the same k value as dilute solutions 

 (Table 6-3), and the absorption curve of such concentrated solutions meas- 

 ured in the microspectrophotometer matches very closely that obtained on 

 dilute solutions (Fig. 6-10). Nevertheless, it is difficult to escape a sus- 



2400 



3000 



2600 2800 



WAVE LENGTH, A 



Fig. 6-10. Absorption spectrum of a 

 2.5 per cent ribonucleic acid solution, 

 obtained (1) in a cellophane bag with 

 a microspectrographieal arrangement 

 (soHd circles) and (2) in a 10-/x Scheibe 

 cuvette with a photoelectrical absorp- 

 tion spectrograph according to Warburg- 

 Negelein (open circles). (Redrawn 

 after Thorell, 1947.) 



