68 
Weights of Human Viscera 
From these results we infer, 
(a) That the average " healthy " organs are all lighter than those of the 
average general hospital population, and probably lighter than those of the 
general population as a whole. The weights are still, however, higher than those 
given by Peacock and Reid or by Clendinning. 
(6) In passing from the mixed hospital population to the class of healthy 
organs, we find in every case the absolute variability is reduced, and by very large 
amounts indeed, the variability of the heart by over 50 p.c, and the other organs 
by amounts even 30 to 40 p.c. of their value. 
(c) Relatively, the healthy spleen is still the most variable organ, and the 
heart comes second, but the kidneys are now close to the heart and the liver not 
very far behind. Disease appears to affect the weights of heart and spleen most, 
of livers and kidneys least. 
We notice that our value for the coefficient of variation of the healthy heart is 
now 17'7 as against the result deduced from Peacock and Reid's measurement of 
19"8 and for the liver 14-'8 as against their 14*3*. We could hardly have antici- 
pated such good agreement, and it certainly tends to confirm the value of the 
coefficient of variation as a fairly " steady " biometric constant. 
4. Influence of Age on the Weigh ts of the Viscera. 
I have first considered the change in the absolute weight of the adult male 
heart with age, and I have then investigated the influence of age upon one set of 
correlations and variabilities, i.e. those of heart and spleen. 
Table VII. gives the correlation between age and weight of heart in the case of 
health. We deduce the following values of the constants : 
TABLE VII. Relationship of Weight of Healthy Heart to Age. Number 699. 
Mean Heart 11-13 ozs. Standard Deviation 2-015 ozs. 
Mean Age 40-23 years. Standard Deviation 8-500 years. 
Correlation of Weight of Heart with Age = -1363 ± -0250. 
There is thus a distinctly sensible increase of heart weight in health with age, 
the coefficient of correlation is more than five times its probable error. Still the 
correlation is smaller than what we might possibly have anticipated. Calculating 
the regression line we have, if be the probable weight of heart in ozs., and 
A the age in years : 
iTp^ 9-8322 + -0323^. 
Thus the average healthy heart gains about I oz. per ten years. For example 
we have : 
* Peaisou (The Chances of Death, Vol. i. p. 318) gives 20-49 for the coefficient of variation of the 
riyht kidney as deduced from Peacock and Keid's measurements, as against my values of 16-80 for 
healthy and 24-63 for general hospital weights of both kidneys. 
