SECTIONAL 'TRANSACTIONS.—J. 469 
It has also been possible to relate intelligence to legitimacy and to such measurements 
of general health as the height-weight index, and a doctor’s estimate of general 
physical fitness. 
The general conclusion suggested by this work is that, while a slight inherent class- 
difference in intelligence seems to have been demonstrated, at least in one group, 
the statement needs most careful qualifications and limitations. 
Prof. F. C. Bartiert.—Psychological Problems in the Government of Native 
Races. 
Dr, SHepHerD Dawson.—Intelligence and Fertility. 
The application of individual mental tests to a large number of children of 
approximately the same social status, and an inquiry into the size of the family to 
which each belonged, showed a small but significant negative correlation between 
the intelligence of the child and the size of the family. The birth-rate was higher in 
the families represented by the dull children, and, although the fatalities were also 
higher, the number of survivors was greater. The difference would probably be 
greater if a representative sample of the whole population were examined. Although 
intelligence is affected by pathological conditions, yet, if it be an inherited character, 
as is commonly believed, this suggests a possible slow but serious dilution of intellect 
in the population as a whole and the need for a study of the inheritance of mental 
aptitudes. 
Friday, September 25. 
Diviston 1. 
Dr. R. H. Tooutess.—Individual Differences in Phenomenal Regression. 
The amount of ‘ phenomenal regression’ of different subjects, i.e. the degree of 
their tendencies to see the ‘real’ characters of objects rather than the characters 
given by peripheral stimulation, was measured by three tests. These were: (A) a test 
of ‘tendency to constancy of shape’; (B) a test of ‘ tendency to constancy of size’ ; 
and (C) a test of ‘ tendency to constancy of brightness.’ The principal objects of the 
experiment were, first, to discover how universal was the effect and how large were 
the quantitative differences between different individuals ; and, secondly, to find out 
the nature of the quantitative relationship between the different tests. 
Of 129 subjects (including ten artists and teachers of art) none were found who 
did not show phenomenal regression in tests A and B, although a few showed zero or 
negative results in test C. The median tendency was large with a large scatter, 
_ showing that, in general, perception is largely determined by ‘real’ characters, but 
that the extent to which it is differs very much from individual to individual. 
There was a large correlation (about -7) between the results of test A and test B 
for groups homogeneous in age and sex. This remained considerable after the effect 
of partial correlation with intelligence had been eliminated. ‘This shows that there 
is a group factor determining the tendency to phenomenal regression, the strength 
of which in an individual determines how far he sees things in their ‘ real’ shape when 
he is looking at them at different angles, and how far he sees them in their ‘ real’ 
sizes when he is looking at them from different distances. 
Eyidence as to whether the tendency to constancy of brightness was dependent 
on the same factor was inconclusive. The correlation of test C with the other tests 
was low and of doubtful significance. Quantitative treatment of this test was 
complicated by the discovery that two variables contributed to the result; the 
tendency to see the more reflective paper as the brighter even when reflecting less 
light, and the tendency to a difference between the phenomena of ‘ whiteness’ and 
‘brightness.’ This difference was but absent in about one in five subjects. Phe- 
nomenal regression for ‘ whiteness ’ showed fairly high correlation with tests A and B. 
Correlation of phenomenal regression with intelligence was found to be small and 
negative, i.e. the more intelligent subject tends slightly to perceive more nearly to 
the stimulus character of the object. The relationship with age was in the opposite 
direction, the older subjects tending to see more nearly the ‘real’ character of the 
object. There was a significant difference between men and women students, the 
latter having the greater tendency to phenomenal regression. A small group of 
