74 Milton Landowne 



"normal" man, or for that matter even a good sample, then 

 one might cautiously venture to propose that agewise changes 

 in such a sample were manifestations of normal ageing and, 

 therefore, these might constitute the indices and the represen- 

 tations of the normal ageing process. The unfortunate fact is 

 that we use the word "normal" here to mean "without 

 abnormality" or "free of pathology", etc., and not in its 

 alternate statistical sense of "norm", referring to the common 

 model, the standard or the representative type. We may try 

 to obtain our group of "pure" normals by excluding known 

 or testable abnormality, but since we can hardly exclude 

 unknown abnormality, we are unable to apply a criterion of 

 the goodness of our selection, and have no test of purity. If 

 all the subjects in an experiment are represented by 100 per 

 cent, such diseased subjects as may be readily identified 

 might constitute, let us say, about 10 per cent of the popula- 

 tion. This itself might not be too disturbing, but an increasing 

 agewise incidence of this fraction does disturb us. We may 

 deal with this by excluding the members of this group in order 

 to obtain greater uniformity and less interfering disease. But 

 do we succeed in this? Although manifest disease may be 

 readily diagnosed, minimal to moderate disease is sometimes 

 found in higher incidence than can be reliably disclosed by our 

 clinical methods of examination. Fig. 1, from the data of 

 Willius, Smith and Sprague (1933), shows the percentage 

 incidence of coronary arteriosclerosis in a series of over 3,000 

 consecutive male autopsies at the Mayo Clinic. The incidence 

 of marked and of extreme disease is of an order compatible 

 with clinical figures, but only 2 out of 1,000 subjects 60 years 

 or older were judged free of the disease on gross morphological 

 examination of the heart. On this basis, an individual of 60, 

 without coronary arteriosclerosis, is abnormal; i.e., he has a 

 statistically improbable condition — he is unusual. Our task 

 is to determine whether cardiac functional changes are in 

 consequence of or despite coronary sclerosis; to ascertain 

 what effects and interpretation these changes may call for. 

 This is not simple. To illustrate, the electrocardiographic 



