M. HOSNY 
287 
assumption are to take tlie partial coeftlcient between scholars and deathrate for 
constant number of persons per house. 
We find*: p'7,s = + -5336 ± -llO? ; 
similarly y.r^^ = - •0504 + -1543. 
Tliere is thus a slight increase in the relation of scholars to deathrate when 
we take a constant number of persons per house, and it is hard to believe that 
the relation is indirectly due to size of family. The second result shows that 
literacy has no relation to the infantile deathrate. Towns like Alexandria, 
Damietta, Port Said, Ismailia, and to a less extent Suez, with a low infantile 
deathrate have a low education rate, and towns like Cairo, Guizeh, Beni Suef, 
Minia, and Assiut, with high infantile deatlirates have high e(]ucation rates. The 
first towns are on the sea or the canal, the second in the Nile Valley; it is con- 
ceivable that the latter are the more unhealthy for the infant ; it would need 
special local knowledge to explain why education has been most accepted above 
Cairof. There does not, however, seem any relation between ignorance, as 
measured by literacy, and a heavy infantile mortality, nor on the other hand can we 
assert that education and European influence have certainly increased criminality. 
* The value of j.r,..v is+ -0207 ±-1508, and is therefore not significant. 
t It is noteworthy that there is no relation between literacy and number of scholars, i.e. education 
of children does not appear to follow the power to read and write in their parents, that is to say 
if we judge by the averages in towns and not by individuals. 
37—2 
