(_e_) That the per cent. classed as “best” in the various letter groups increased steadily from 0 per cent. in “D−” to 57.7 per cent. in “A,” while the per cent. classed as “poorest” decreased steadily from 80 per cent. in “D−” to 11.5 per cent. in “A.” The following table shows the per cents. for each letter group: ══════════════════════╤══════╤══════╤══════╤══════╤══════╤══════╤══════ │ D− │ D │ C− │ C │ C+ │ B │ A Total number │ 29│ 60│ 121│ 231│ 229│ 191│ 104 Classed with “best” │ 0.0%│ 6.7%│ 19.0%│ 26.0%│ 39.3%│ 53.4%│ 57.7% Classed with “poorest”│ 79.3%│ 66.0%│ 57.9%│ 31.2%│ 24.9%│ 16.7%│ 11.5% ──────────────────────┴──────┴──────┴──────┴──────┴──────┴──────┴────── Considering that low military value may be caused by many things besides inferior intelligence, the above findings are very significant. 2. In an infantry regiment of another camp were 765 men (Regulars) who had been with their officers for several months. The company commanders were asked to rate these men as 1, 2, 3, 4, or 5 according to “practical soldier value,” “1” being highest, and “5” lowest. The men were then tested, with the following results: (_a_) Of 76 men who earned the grade A or B, none was rated “5” and only 9 were rated “3” or “4.” (_b_) Of 238 “D” and “D−” men, only one received the rating “1” and only 7 received a rating of “2.” (_c_) Psychological ratings and ratings by company commanders were identical in 49.5 per cent. of all cases. There was agreement within one step in 88.4 per cent. of cases, and disagreement of more than two steps in only 7/10 of 1 per cent. of cases. 3. In another camp the company officers of a regiment were asked to designate the ten “best” and ten “poorest” privates in each company. The officers had been with their men long enough to know them thoroughly. Comparison of the officers’ estimates with the results of intelligence tests brought out the following facts: (_a_) Of 156 men classed with the ten “best” in their respective companies, only 9 tested below C−. (_b_) Of 133 men classed with the “poorest” ten in their respective companies, only 4 tested above C+. (_c_) Men above C+ are 7.3 times as likely as men below C− to be classed with the ten “best.” (_d_) Men below C− are 10.8 times as likely as men above C+ to be classed with the ten “poorest.” (_e_) An “A” man is 11.7 times as likely as a man below C− to be rated “best”; but a man below C− is 13.5 times as likely as an “A” to be rated “poorest.”