Miscellanea 
371 
Boundaries between Categories and Means of Categories vieasured from 
the Mean of Intelligence. 
Quick Intelligent and Intelligent 
104'5294 rnentaces or 
16-7602 i.Q.u.'s. 
Intelligent and Slow Intelligent 
■4-5294 
•7262 „ 
Slow Intelligent and Slow 
- 77-7640 
- 12-4686 „ 
Slow and Slow Dull 
-141-1658 
- 22-6345 „ 
Slow Dull and Very Dull 
-200-6975 „ 
-32-1797 „ 
Mean of Quick Intelligent 
152-967 
24-527 
„ Intelligent 
49-756 „ , „ ^ 
7-978 
,, Slow Intelligent 
- 34-410 „ „ ' 
- 5-517 
„ Slow 
- 105-552 „ 
- 16-924 „ 
„ Slow Dull 
-165-590 
-26-551 „ 
„ Very Dull 
-235-299 
-37-728 
After careful consideration (if a number of factors we divided-* our " Intelligent" category of 
100 mentace.s range into "Fair intelligence" for the first 45 rnentaces and "Capable" for the 
remaining 55 rnentaces. Our "Quick Intelligent" category was again subdivided into a i-ange 
of 200 rnentaces corresponding to "Specially Able" and to "Genius" or the 1-4 per mille who 
exceed the mental type by more than 300 rnentaces. The "Very Dull" were again subdivided 
at 300 rnentaces less than the mean and the 1-4 per mille lieyond this may be looked upon as 
mentally defective. This per mille of mental defectives corres})onds fairly well with the primary 
school returns. Thus the avei-age 'genius' will have 312 mentaces or be almost exactly 50 
I.Q.u.'s above mediocrity, i.e. with a mean of 143 i.Q.u.'s, and the average mentally defective 
312 mentaces or 50 i.Q.u.'s below the type, i.e. will have about 43 instead of 93 for intelligence 
quotient +. These limits are marked on our diagram. 
Dr Gordon's results therefore bring out a point that was not correct in my diagram of 1906. 
The zero of intelligence is not aliout 300 mentaces below mediocrity, but nearer 600 ! Even an 
"imbecile" girl has an intelligence quotient of 29, or some 180 mentaces, where I in 1906 
assumed she should be credited with none. I still think complnte imbecility should be marked 
by a total absence of mentaces or by a zero intelligence quotient. It apjaears better therefore to 
talk of those with intelligence less by 300 mentaces than the mean as mental defectives J. The 
problem is rather theoretical than practical, depending not so much on the existence of zero 
intelligence, as on the limen or threshold value at which we are alile to realise its existence. 
Anyhow the conclusion seems to be that we must search a large munber of millions if we wish 
to find an individual absolutely without intelligence. 
Examining our diagram we note how extremely closely the black points which represent the 
means of the general intelligence categories lie on their regression line. They lie so closely that 
we might almost feel disappointed that the means for the Slow Dull and Very Dull categories 
are not equally close to the regression line. But here regard must be paid to the fact that these 
are the smallest of the categories in size; and further to disturbing factors arising from the 
* See Bionietrika , Vol. v. p. 110. 
t Dr Gordon notes a very able girl with 137 i.Q.u.'s and an imbecile girl with only 29 i.Q.u.'s in 
a total of 335 cases. 
+ I wrote in 1906 (Iliometrika, Vol. v. p. Ill ft.) that: "He [the median individual] can hardly 
have more than 350 to 400 mentaces, for at a negative position of - 360 to - 400 on the scale we have 
passed through the very dull group into imbecility and complete absence of reasoning power. The 
child whose low grade of intelligence occurs only 3 or 4 times in 100,000 cases must be sought in the 
idiot asylum." I was probably wrong in assuming the worst type of idiot had zero intelligence. 
Dr Gordon's mean is 6 06 times her s.d., or the absolute zero of intelHgence would only occur 1 in 
100,000,000. This is probably excessive. Dr Jaederholm's data appear to indicate 5-5 times as the 
ratio or 1 in 12,500,000 as the occurrenoe. 
