Figure 6.

The left-hand panel indicates that enterocytes do not lie in an orthogonally arranged grid pattern on the basement membrane (plan view, upper diagram). Therefore, “counts” of enterocytes (or more importantly their nuclei) cannot be accomplished with the ease often assumed in the Methods sections of many publications. This model obviously predicts the possibility of observing large tracts of enterocytes without nuclei (as in the imagined sections at (A,B)), an event never encountered in histopathological practice. Only occasionally would a palisade that included a run of every adjacent enterocyte, and their contained nuclei, be observable (C). Therefore, this model is wrong. The alternative (right-hand panel) is closer to reality, comprising an idealised epithelium, scaled to data obtained by transmission/scanning EM studies. The lines in the upper (plan) diagram reflect random sectioning planes through this epithelium. But, it should be carefully noted that, on average, only ~50% nuclear profile discs appear in any section, as represented imaginatively in A,B,C below. Thus, the high numbers of “lost” enterocyte nuclei now becomes apparent. However, since IEL counts are made relative to the simultaneously changing world of enterocyte (nuclei) populations, values are spuriously increased twofold. The basic flaw is discussed elsewhere (reference [55]: and see Figure 5).