W. H. Gilby 
97 
are several fallacies that may arise in the interpretation of this correlation. It 
might be asserted that the more intelligent child will look after his clothing better, 
but although this may produce some effect on general tidy appearance, it cannot 
contribute largely to influencing an estimate based on the existence of sufficient 
or insufficient underclothing and the presence of boots, etc. There can we think 
be little doubt that the evidence of clothing is roughly a measure of home 
conditions. But it is none the less fallacious to assert that the inferior intelligence 
evidenced by the poorer clothing is necessarily a product of bad home environ- 
ment. It will be clear that intelligent parents who have usually higher wages 
will provide better clothing and will look after their children better. Hence the 
problem turns — as most such problems do — on the relative intensities of nature 
and nurture. Is the lower intelligence of the children due to the poorer home 
environment evidenced by the worse clothing, or is the worse clothing only a 
mark of the lower intelligence of the parents, which is naturally reproduced in 
their children ? If we look into this point algebraically, we may write the 
subscripts 1 = intelligence of the children, 2 = intelligence of the parents, 
3 = clothing of the children. Then what we really want is: 
/y /y 
13 ' 12 < 23 
i.e. the correlation between clothing and intelligence of the children for constant 
intelligence of the parents. If this correlation be substantial, then the proposition 
that the intelligence of the children is directly influenced by their environment 
will receive some support. The problem then turns on whether r 12 x r 23 is of 
sensibly the same order or not as r 13 . We may safely say that r 12 lies between 
"4 and *5 ; what value are we to give to r 23 ? No direct estimate of the relation 
of parental intelligence to the clothing of the offspring is at present available, 
but we think few would be hardy enough to assert that it would be unreasonable 
to consider it as lying between "4 and '6 at the least. The commonest experience 
seems to show that a tidy child followed home will disclose a careful intelligent 
mother, and a father, whose intelligence is measured by adequate wages. Until, 
however, this point has been definitely examined statistically, it is futile to 
dogmatise about clothing being a standard of parental neglect, and that such 
neglect is producing poor intelligence in the offspring. A better argument might 
be deduced, if it could be shown that adequately fed and clothed pauper and 
asylum children are of superior intelligence to the children of the public primary 
schools. Many other tests of the presumed influence of environment are of the 
like superficial character to the clothing test as evidence of the influence of home 
environment ; they are fallacious until they have been modified by correction for 
the hereditary factor*. 
* It is not unusual for the school medical officer to find a correlation between intelligence and 
evidences of parental neglect, dirtiness, poor clothing or inadequate nutrition. It does not follow that 
this relationship which lies between -2 and -3 is the source in whole or even part of the poor intelligence. 
It may be, but the evidence given, which wholly neglects the hereditary factor, is quite insufficient to 
prove that it is. 
Biometrika vin 13 
