A CRITIQUE OF CYTOCHEMICAL METHODS 211 



entities by color reactions. Within the dimensions of the cell, diffusion — 

 11 slow process at the macroscopic lev^el — is practically instantaneous, and 

 there is every reason to suppose that extensive redistribution takes place, 

 either as tissue is being fixed or as the test is carried out on the tissue. 

 This is particularly misleading because the precipitates tend to be 

 adsorbed on the extensive internal surfaces of the denser parts of the cell 

 such as nuclei, myofibrillae, or the thick distal borders of some epithelial 

 cells. Thus many early observers erroneously reported that the nucleus 

 contained considerable iron, an element in which it is actually notably 

 deficient. The nucleus was repeatedly described as the site of the enzyme 

 alkaline phosphatase, since phosphate split off from a substrate, glycero- 

 phosphate, appeared as an intranuclear precipitate, but Jacoby and 

 Martin (1949) have demonstrated that this is largely a secondary accumu- 

 lation of the phosphate (see also Novikoff, 1951, 1952). 



Redistribution during the fixation process is effectively prevented by 

 the freeze-dry method of preparing sections of tissue, which was suggested 

 by Altmann (1890) and elaborated by Gersh (1932), Hoerr (1936), Simp- 

 son (1941), and others. In this technique fresh tissue is quickly frozen 

 at such low temperatures ( — 190°C) that ice crystals do not form (the 

 water appears rather to be practically ^^itrified). Tissue is then dehy- 

 drated at low pressure and temperature and is finally embedded in paraf- 

 fin and sectioned. Up to this stage, it is generally agreed that little 

 redistribution of intracellular chemical constituents can have taken place, 

 and two excellent methods of elementary chemical analysis apparently 

 can be carried out without producing any essential change in this dis- 

 tribution. The paraffin sections may be burned in an electric furnace 

 (microincinerated; Policard, 1923; Scott, 1943), and the appearance and 

 amount of the ash indicate the distribution of the mineral elements in the 

 cells (e.g., iron is a yellow to red ash, silicates are crystalline, and calcium 

 and magnesium are amorphous and dense white ashes.) In another 

 method the paraffin is removed from the section, and the tissue is dried 

 and subjected in vacuo to X-ray absorption analysis (Engstrom, 1946, 

 1950). Although it involves immense technical difficulties, the latter 

 appears to be an extremely promising approach since it offers the possi- 

 bility both of determination of cell mass from polychromatic X-ray 

 absorption data and of analysis for a large number of individual elements 

 from absorption of monochromatic X rays in the wave length range 

 2-50 A. 



The distribution in frozen-dried sections would also be expected to 

 remain essentially unchanged, except for lipoidal constituents, if the 

 paraffin is removed — for instance, by chloroform — and the section kept 

 in nonaqueous solvents such as alcohol and glycerin. This procedure has 

 been recommended in preparing material for ultraviolet absorption 

 studies of intracellular proteins (Caspersson, 1947) principally because 



