234 RADIATION mOLOGY 



give a markt'cl plasmal reaction with SchifT's rcaf^ont. The structures 

 may be isolated by crushing or smearing the cell, the method used l)y 

 Caspersson (H)3()) with components of the salivary gland nucleus and by 

 Ris and Mirsky (1949) with nuclei. This is, however, a procedure by no 

 means always applicable. The problem is probably best solved by sec- 

 tioning tissue approj)riately. Nuclear and luicleolar absorption are thus 

 easily isolated if sections are cut approximately ecjual to or le.ss than the 

 respective diameters of these cell components. The techni(|ue of sec- 

 tioning material in a much lower range of thickness has been developed 

 for electron microscopy (Hillier and Gettner, 1950); sections can be cut 

 1 ^t or less, and thus the absorption of single mitochondria or secretory 

 granules can possibly bo studied without interference from cytoplasmic 

 overlay or underlay. Until now mitochondria have been studied only 

 where a considerable number are fused to form a large body, the neben- 

 kern of the insect spermatid (Leuchtenberger and Schrader, 1950). A 

 similar advantageous concentration of small bodies (e.g., microsomes) 

 can be accomplished by ultracentrifuging (Lagerstadt, 1949). 



In studies of living cells, this problem of isolating the absorption of a 

 particular component seems nearly insolvable. Rarely, if ever, can it 

 be arranged that the nucleus is relatively free of overlying or underlying 

 cytoplasm; either this must be ignored (Caspersson, 1939; Thorell, 1947; 

 Mellors et al., 1950) or a correction must be attempted by measuring the 

 c3''toplasm, estimating the thickness above and below the nucleus, and 

 subtracting the proportional cytoplasmic absorption, assuming that it is 

 homogeneous (Caspersson, 193(5). This assumption is likely to be incor- 

 rect since in many cells the ultraviolet absorbing material is especially 

 concentrated just outside the nuclear membrane. Much more difficult is 

 any sort of measurement of absorption of a nucleolus, and it seems inad- 

 visable to attach much significance to the apparently uncorrected absorp- 

 tions of small nucleoli in living cells (Thorell, 1947). 



The problem of fixation for chemical cytology is of major importance, 

 even though it has not received much attention, cytochemistry having 

 merely taken over some orthodox cytological technicjues. The.se are by 

 no means all suitable methods of observing tissue for cytochemical study. 

 For example, the best fixation of the complete structural details of cells 

 is accomplished by use of mixtures which include osmium tetroxide 

 (osmic acid, Os2()4) and chromium trioxide (chromic acid, CrOs), (Fig. 

 6-2^4), but these fluids are of extremely limited application in cytochem- 

 istry. The chrom-osmium mixtures always tend to overemphasize phase 

 boundaries, they introduce color into the section, and, after their use, 

 practically no fractions can be removed chemically or enzymatically. 

 C'ytochemical fixation, as a rule, has been a compromise between the 

 requirements of accurate preservation of morphology and tho.se of photo- 

 metric analysis and fractionation (Pollister, 1952a). Two colorless 



