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



closer range) is due to the same difference of general illumination. 

 We are then, it is true, left at loose ends in the matter of R. Here, 

 if comparson is at all permissible, Geissler's results are in agreement 

 with Rivers' ; my own are in disagreement. But we do not know 

 whether the two R glasses were alike. Above all, I cannot tell 

 whether the tintometer-series of colored glasses are so combined 

 with grays as to show the same " brightness " throughout, or whether 

 during an experiment they brighten with decrease of chroma ; I can- 

 not tell, either, whether the colorless field, with which the colored field 

 is compared, itself varies with the " brightness " of the colored glasses 

 or remains the same for all.^^ In my own experiments " brightness " 

 varied with chroma, but the colored and colorless fields were of ap- 

 proximately the same "brightness" in every observation. Here are 

 possible differences of procedure that might affect the results. 



I said just now that we are not informed of the conditions under 

 which the English tests were made. It is worth noting, however, 

 that Rivers found excessively high limens " in testing Europeans in 

 too strong a light."-" Since too strong a light (daylight) could 

 hardly be obtained save in a specially lighted laboratory room, it 

 seems probable that my supposition as regards the placing of the 

 tintometer is correct. 



All of these criticisms are offered, of course, with the greatest 

 reserve ; Rivers may be able to meet them point for point. Taking 

 his report as it stands, I think they are sufficient to cast serious 

 doubt upon the conclusion which he draws from the tintometer- 

 results. The report, however, is incomplete; we lack details of ex- 

 perimental procedure and conditions ; and the English observers, who 



25 Lovibond speaks of neutral-tint glasses, standards, units ; he also has 

 dififusive glasses, thin slips of ground glass (op. cit., 21, 31 fif., 48, loi f.). It 

 would therefore seem possible to keep hue and " brightness " of the colored 

 glasses constant while chroma varied, and to equalize the " brightness " of 

 the colored and colorless fields. Rivers does not tell us whether this was 

 done. In setting up the instrument for differential determinations, he found 

 that the " difference in brightness " of the glasses rendered the results incon- 

 clusive (R, 74). If brightness affected these results, must it not have affected 

 the others? and if it was compensated in the other experiments, might it not 

 have been compensated in these? 



2« R, 72- 



