482 A Second Study of the Human Viscera 
excepting the variability of the heart, no appreciable alteration can be discerned. 
Such changes as there are, e.g. the increased variability of the heart and its 
lessened correlation with the other organs, point to the first selection as being 
somewhat more healthy than the whole series, but the differences are evanescent. 
The regression equation of heart upon body and kidney weight as determined from 
the whole series is : 
x 1 = -0462^ + -4260«3 + 1-6143 (4), 
where x x = heart weight in ounces, x, 2 = body-weight in pounds, x s = kidney weight 
in ounces. 
The mean error for the 26 test cases is 1*1 ounces. 
We think the inference to be drawn from this comparison is that, in order to 
arrive at the values of the true post mortem constants descriptive of a healthy 
" General Hospital Population," the process of selection need not be very stringent, 
so far as the diagnosed cause of death is concerned. 
We also think the whole tendency of the present investigation is to suggest 
that the constants here published are, within the limits indicated by the errors of 
sampling, representative of a healthy " General Hospital Population." None of 
our results are in flagrant opposition either to the findings of observers equipped 
with more accurate data or to those previously obtained from hospital series. We 
hope, therefore, to have added another item of use to the student of a " General 
Hospital Population." Of course, being restricted to a single type of measurement, 
this study is of but limited interest. Shrubsall has published* the results of an 
inquiry into certain anthropological characters of a " General Hospital Population," 
while Greenwood and Candy f have shown cause for believing that the fatality of 
certain common diseases, such as lobar pneumonia, may be quite different in this 
class from that obtaining in the community at large. These are other items in 
a catalogue which has yet to be completed. It may well be possible for a future 
synthesist to paint a fairly accurate composite picture of a section of our country- 
men which is of immense importance not only in itself but also as the substratum 
of medical instruction. At present, however, we must be contented with the task 
of collecting material, in particular statistical records. 
The general conclusions of this inquiry may be summarised as follows : 
(1) The mean weight of the heart in healthy adult males of the class from 
which patients in a London General Hospital are drawn is not far short of 
12 ounces. 
(2) The correlation between the weight of the heart and that of the whole 
body is not much less than - 5. 
* F. C. Shrubsall, Brit. Med. Journ. Vol. n. p. 1689, 1904 ; St Barts. Hosp. Reports, Vol. xxxix. 
p. 64. 
t Journ. Roy. Statist. Soc. Vol. lxxiv. p. 365, 1911. 
