682 TRANSACTIONS OF SUB-SECTION I. 
the middle, and that later it rises again. Experiments have been carried out 
to test this with the following results :— 
Tasie I. 
Average Changes in the Spatial Threshold on Right Forearm during the 
Progress of a Sitting. 
Subject No. 6 Subject No. 7 Bublect ee Wetenied 
ems. | ems. | ems. | ems. | ems. | ems. | ems. cms. cms. 
i| 2-98 | 2-15 | 3-25 | 3-30 | 3-28 | 3:99 | 4-50 3:50 | 3°38 
i ii] 2-66 | 1:95 | 3-11 | 2-95 | 2-28 | 3-29 | 4-57 3:20 3:05 
=| iii} 2-40 | 1-85 | 3-08 | 3-25 | 2-68 | 3:34 | 4-42 2-60 2:97 | 
s iv| 2-16 | 2-19 | 2-58 | 3:05 | 3-28 | 2:99 | 4-17 | 2-45 2-83 
el v/ 1:84 | 1:59 | 3-16 | 3-40 | 3-28 | 2-84 4:80 2-85 2-95 
3 vi| 2-01 | 1-72 | 2-29 | 2-95 | 3:08 | 3:09 | 3-97 | 2-95 2-74 
~ | vii} 1-95 | 2-05 | 2-61 | 3:35 | 3-68 | 3-59 | 4-95 | 3-20 3:14 
© | vili| 2:06 | 2:05 | 3:06 | 2:45 | 3-88 | 3-34 412 | 3-05 2-94 
S | ix] 2-52 | 1-92 | 2-68 | 3-05 | 3-68 | 3-44 3°75 3:15 2-98 
a | x 2:06 | 1:85 | 2-15 | 2-65 3-28 | 3-44 4-45 3:05 2-84 | 
Each of these columns is based on ten sittings, each of 100 judgments and 
lasting about twelve minutes, except column 5 based on five sittings only. Each 
sitting is divided into ten periods, shown by Roman numerals. The first figure, 
for example, 2.98 cms., is calculated from 100 judgments—namely, the first ten 
judgments at each of ten sittings. For subject 6 only, the sitting proper was 
preceded by a preliminary group of ten judgments. With this subject the 
Method of Non-Consecutive Groups” was used, and the results were calculated 
by Urban’s formula.* The method used in the other cases would be defined, in 
the nomenclature suggested by the writer,* as the Method of Right and Wrong 
Cases with catches, using the direct Limiting Process of calculation. 
The results suggest that often the threshold falls sharply during the first 
twenty judgments and then slowly up to about forty judgments, after which it 
becomes erratic, but often falls again at the very end of the sitting. The 
practical advice would be to give preliminary judgments and not to prolong 
the sitting. Space does not permit a discussion of whether these results are 
‘significant ’ or due to chance sampling, but this will be done elsewhere. 
Each of the numbers given above is itself an average, and the variation is 
much greater than they indicate. Whether we look upon this variation as due 
to experimental error, or as a true variation® depends upon our point of view. 
There is no essential difference. 
Taste IT. 
Spatial Threshold on Right Forearm. Subject No. 6. Method of 
Non-consecutive Groups. 
Stimulus value in Gms. | 0 0-5| 1 | 15} 2 25 3 /35| 4 j45| 5 | 
| No. of answers two 52 | 6 | 8 | 21 | 56 | 84 | 105 | 125/141 | 144 | 148 
| | 
| Total No. of Judgments | 1500 150] 150] 150 | 150) 150) 150) 150] 150) 150] 150 
An average threshold is a matter of mathematical definition, not of psycholo- 
gical experience. It has been defined as the point where 50 per cent. of correct 
? Thomson, Brit. Journ. Psychol. 1912, v. 204, 214. 
° Archiv. f. d. ges. Psychol. 1909, xv. 291, 307. 
* Op. cit. 204, 210, 240. 
Cf. Brown, Mental Measurement, Camb, 1911, 20 footnote (references 
to Miller, Fechner, Bruns) and 84. 
5 
Prcew is' 
