1896-97-] SPATIAL THRESHOLDS OF COLOURS. 233 



precautions. And, as a matter of fact, the divergence is quite according 

 to the expectations we would naturally form in regard to the matter. In 

 fact, so materially and in all points different were the conditions under 

 which these investigations were carried out that it is scarcely instructive, 

 scarcely worth while to institute any except the briefest comparison 

 between their results. 



However, before referring further to the results of a more accurate 

 method, we will briefly notice a set of experiments which we made upon 

 this problem, largely under the same crude conditions as Von Wittich 

 and Aubert made theirs. We employed, as they did, uncorrected pig- 

 ment papers for our colours, and daylight illumination. We had, how- 

 ever, a very fine specimen of black ground, approaching about as close 

 as possible to an absolute black — which we attained, not by using a 

 black cardboard illuminated by the same light as the colours, but a 

 black surface from which all daylight and every other kind of light was 

 rigorously excluded, so far as possible, by means of diaphragms. It 

 was on such a black ground that our colours were viewed. With the 

 exception of this difference, and the fact that we employed the accurate 

 measurement apparatus referred to before, our results in this particular 

 set of experiments were obtained under very much the same general 

 conditions as Aubert's and Von Wittich's. However, we find, as we 

 would expect from the impossibility of the actual duplication of the 

 conditions of colour mixture in the pigment colours, and the exact 

 reproduction of the precise contrast influences at work in the experi- 

 mentation of Aubert and Von Wittich, that our results differ quite con- 

 siderably from theirs, just as theirs mutually differ. In the main our 

 results show a lower achromatic threshold than Von Wittich's, and 

 generally quite as low as, or, with a few exceptions, even lower than the 

 findings of Aubert. But the case is different with the chromatic and 

 characteristic colour thresholds, in which we are somewhat higher than 

 Von Wittich and Aubert. For example, Aubert's light thresholds on 

 black ground vary from i' 14" blue (ultramarine) down to 39" orange, 

 and red at 59" which are rather low ; while Von Wittich's light thresh- 

 olds on the black ground are never lower than i'4" and though in that 

 neighbourhood for the most part, yet, in the case of violet, reach as high 

 as 3' 26". On the other hand our similar results show a fairly uniform 

 tendency to have the light thresholds within the region between 46" and 

 l' with the exception of yellow and grey which reach i' 14" and i' 17" 

 respectively. In regard to chromatic and characteristic thresholds Aubert 

 has some very low, such as orange and red at 59" and others that go up 

 as high as 3' 47" and 4' 17", rose and blue respectively ; while Von 



