TRANSACTIONS OF SUB-SECTION I. 681 
The chief deterrent factors are that no records of childless families are 
possible, and that the smaller families were not perhaps complete, although in 
over 80 per cent. of cases this was the case. 
Correlation between the size of the fraternity and the number of members 
who had died :— 
No. Size of Number 
Type Observed eet A Dead $ - 
Defective. .| 1,000 646 | 3:35 1:91 2-15 ‘74-01 
Normal . ./| 3,000 4:59 | 2°48 0-48 0:89 +5 + 009 
This gives an average net family of 4°55 for the defective and 4°11 for the 
normal, but by the time the central child has reached adult age the disparity 
would probably have disappeared. 
From a study of 700 families it appeared that on an average 1°43 children 
were mentally defective in a fraternity of the average size of 6°05. The known 
number of defective members in each family was, however, much higher, being :— 
No. a No. of Families No. Affected No. of Families 
6 10 
2 148 a 6 
3 719 8 5 
4 49 9 2 
5 32 10 Li 
11 1 
The fact of mental deficiency was found to have been noted :— 
In infancy in aoe ee ee GL per cent. of cases 
By the age of four Se Fiat cs outs tes 2 ay ds er Cent: Olreases 
By the age of five - «ss. » .« .2°5 per cent, of cases 
On going to school - .« . «  « 28°5 per cent. of cases 
5). The Relation between Habit and Memory. 
By Miss May Smiru. 
This paper dealt with 
1. Bergson’s two forms of memory which he considers to be fundamentally 
distinct. 
2. Experiments illustrative of (a) pure memory; (b) habit interpreted by 
memory. If the two forms are distinct, then tests belonging to each 
group should show high correlation with the others of the same group, 
but low correlation with those of the other group. 
3. The results of the experiments worked out according to the method of 
correlation. 
4. General considerations. 
6. On Changes in the Spatial Threshold during a Sitting, and on the 
Nature of Thresholds in General, By Gopvrrey H. THomson. 
The writer has previously’ suggested that during a sitting of one hundred 
judgments the spatial threshold is frequently higher at the beginning than in 
Brit. Journ. Psychol, 1912, v. 233. 
