126 PSYCHOLOGICAL METHODS OF TESTING INTELLIGENCE 



takes the entirely different problem of estimating 

 their intelligence. 



With the teacher in the higher schools the situa- 

 tion is different. He instructs only in certain sub- 

 jects and thus comes to know his pupils only par- 

 tially. This certainly renders the estimation of their 

 intelligence difficult. The departmental teacher 

 must especially guard against identifying special 

 talent or lack of talent in his special subject with 

 general intelligence or the lack of it. Yet he is less 

 biassed in his construction of a rank-order by the 

 same circumstances, in that the rank-order of the 

 school, if there be one, is never his own work that 

 might affect his judgment by auto-suggestion, but is 

 something that has been obtained by the mere me- 

 chanical addition of all the different performances 

 of the pupils, including performances in other sub- 

 jects with which he has nothing to do. And this en- 

 tails a further advantage, viz., that estimations of 

 the intelligence of the same group of pupils can be 

 obtained from the different departmental teachers 

 who instruct them, and that these estimations can 

 then be compared with each other and finally amal- 

 gamated into a single series. Of course, for the pur- 

 pose of making a comparison of this sort only those 

 teachers should be drawn upon whose subjects of in- 

 struction can afford a basis for a fairly exact knowl- 

 edge of the pupils not, then, some teacher who 

 might give instruction to the class in some minor ac- 

 cessory subject only. 



It is evident, also, that teachers can be asked to 

 make estimations of their pupils' intelligence only 

 after they have become intimate with them not, for 



