108 TITCHENER AND PYLE— JUDGMENT OF DISTANCE. [April i8, 



toward the last than the first." ^ For this reason, " in computing 

 the averages, those for the shadowless series which were taken at 

 the same time as the series with the illusion ' long ' were kept 

 separate from the averages for the shadowless series taken at the 

 same time as the series in which the illusion was * short.' " ^ It is 

 clear, however, that shadowless and shadow-series could not, in 

 strictness, be taken at the same time; they were taken successively. 

 And, as the change of judgment was progressive, Dunlap's averages 

 are used for the comparison of results that are, in strictness, in- 

 comparable. Under these circumstances, it is entirely possible that 

 chance, in determining the order of the single series, may have 

 played, so to say, into the hands of the illusion motive. (4) Dunlap 

 does not tell us how he measured his lines : whether behind the 

 screen, from angle to angle, or in front of the screen, from marker 

 to marker. If he measured behind the screen, then the movement 

 of the right-hand angle only every fifth time that the right-hand 

 marker was moved ^ would introduce a constant error, which must, 

 necessarily, operate in the same direction as an effective illusion 

 motive. (5) Lastly, it may be observed, in general,, that observers 

 in method-work, however well-meaning, fall easily into a reliance 

 upon secondary criteria ; and that an apparatus of the kind used by 

 Dunlap might easily admit this source of error. This suggestion 

 must remain vague, since, without actual trial of the apparatus, 

 we cannot say what the nature of the secondary criteria would be; 

 the suggestion itself, however, does not seem to us unfair, whether 

 in the light of our own experience or in that of Dunlap's account 

 of his procedure. 



To attempt, in this manner, to explain away the results obtained 

 and the conclusions offered by another investigator is not a grateful 

 task. Some of our suggestions may be put out of court at once 

 by a word of explanation from Dunlap. The suggestion of a pos- 

 sible miscalculation — made by us, be it repeated, only on the ground 

 of positive evidence of careless treatment — should be offset by the 

 admission that Dunlap planned his experiments carefully, and with 



'^Op. cit., 447 f. 

 ^Op. cit., 448. 

 ^Op. cit., 442. 



