- CELL COUNTING 425 



However, the fresh-frozen sections were fixed and stained before 

 the counting so that a correction is required to compensate for the 

 contraction produced by this treatment. Since the counting was car- 

 ried out through the entire thickness of the section, the contraction 

 in thickness need not be considered because the ratio of the actual 

 number of cells to the cells counted will not be altered by a change 

 in thickness. The change in area must be considered, however. The 

 ratio of the area of each stained section to that of the fresh section 

 (5.47 mm.-) was obtained by projecting both images at the same 

 magnification and measuring the respective areas with a planimeter. 

 The areas might also be obtained by tracing the outlines of the pro- 

 jected images on paper and weighing the paper. Then the actual 

 number of cells per section is 



/No. stained cells \ ^ /no. counting areas \ ^ / area of stained section X 

 V per counting area/ \per fresh section/ \ area of fresh section / 



And this quantity divided by the volume of a fresh section equals 

 the actual number of cells per unit volume of fresh tissue. 

 In the present case: 



/5.08\ 



(163 ^ 3.1) (85.7) V5.47; ^ (igg ^ 3 7) ^q, eells^l. 

 0.066 



In regions of inhomogeneity within the section the relative areas 

 of the mutually deviating regions were measured after projecting 

 the images on paper, and the separate regions were counted. Each of 

 the regions of a given cell type was made to contribute to the final 

 mean result as illustrated in the following examples from Rask- 

 Nielsen (1944): 



(l) In the case of a section containing one type of cell not homogeneously 

 distributed: 



In 61% of the section area the cells were arranged in glandular tubules. Mean 

 of five random counts was 306. Mean error ±6.3. 



In 39% of the section area the cells were arranged in pits. Mean of four 

 random counts was 203. Mean error ±14.8. Mean number of nuclei plus frag- 

 ments per counting area was: 



(306 ± 6.3) 0.61 + (203 ^ 14.8) 0.39 = (187 ± 3.8) -f- (79 ± 5.8) = 266 ± 9.6 



Applying the correction (0.71) for nuclei fragments: 



(266 ± 9.6) 0.71 = 189 ± 6.8 cells /counting area 



and then following the formula used previously: 



