E. B. TITCHENER— ETHNOLOGICAL TESTS OF SENSATION. 213 



on the " discrimination of dual cutaneous impressions."^* They 

 have not been pubHshed; but by the kindness of my colleague, Pro- 

 fessor Kennedy-Fraser, I am allowed to quote from them in this 

 place. The aim of this test also is to " seek a distance such that 

 about 8 correct judgments in lo are made, i. e., such that double con- 

 tact is reported as double in 8 of lo trials." The condensed results 

 from eleven subjects, obtained after the prescribed preliminary 

 practice, are given in the table on page 212. 



The subjects of this test received no instruction regarding per- 

 ceptive form, and the range of the limens (25 to 52 mm.) may be 

 partly due to that omission. If we consider the results solely from 

 the quantitative side, we may draw the following inferences: 



1. It is dangerous to repeat a test-series (Subject 2) ; to repeat 

 is to give the method an opportunity to contradict itself. Conversely, 

 the result comes out most neatly with the use of few test-series and 

 fairly wide steps. 



2. Inversions are not uncommon. Subjects i, 2, 4 and 6 furnish 

 instances in which a lesser separation of the compass-points yields a 

 greater number of correct judgments. 



3. The weight to be attached to the preliminary trials is left to 

 the discretion of the experimenter (Subjects i, 10). The test-pro- 

 cedure thus contains an element of uncertainty. 



4. The test as prescribed may break down altogether (Subject 

 7). The failure in this particular instance is due, not to the inter- 

 currence of paradoxical judgments,^^ but to irregularity of the nor- 

 mal judgments; the subject evidently changed his standard as the 

 trials w'ent on. 



1^ G. M. Whipple, " Manual of Mental and Physical Tests," 1910, 207 fif. 



1" This possibility was foreseen by Whipple (op. cit, 211). The discard- 

 ing of series by the experimenter (Subjects 4 and 11) was done for cause; 

 but it too introduces an element of uncertainty. 



To meet the variety of perceptive form, Whipple recommends in his sec- 

 ond edition (I., 1914, 247 ff.) a method of contrast. "The threshold may be 

 taken as the distance at which two errors are first made with the ten double 

 points, unless subsequent better records with lesser separations show that 

 these errors were due to a temporary lapse of attention." I am afraid that 

 repetition would still be dangerous ; not only attention but also basis of 

 judgment might shift from series to series. 



