Intelligence and Miscellaneous Articles. 381 



disk so as to employ the previously unused surfaces. Small irre- 

 gularities of texture or a sandy appearance do not interfere with 

 the use of the disks to a noticeable extent, but when the depth of 

 the tint changes slowly over an extended surface it gives more 

 trouble. Of course the case frequently arose where in the entire 

 hundred disks not one could be found with which the flickering 

 entirely disappeared at quite low rates of rotation, and here it 

 was necessary to determine with which grey disk the flicker was at 

 a minimum. This was accomplished by combining the same coloured 

 disk with two grey disks, a larger and a smaller, on the same axis, 

 when it would become evident which was the more favourable 

 combination, and the observer would notice that the luminosity of 

 the coloured disk must be nearer to one grey disk than to its mate, 

 or about equally distant from both. 



As before stated, results w T ere obtained for six disks, but these 

 were selected so as to be complementary to each other in pairs, 

 and in order to test the process they were now combined pair-wise 

 and the resultant luminosities of their grey mixtures were deter- 

 mined by the old method, and afterwards calculated on the basis of 

 the figures furnished by the flickering process : — 





Grey mixture. 



Difference. 





J 27'5 observed. 

 \ 27*5 calculated. 





Purple and green . . 







Eed and blue-green 



1 20*2 observed. 

 1 21 '1 calculated. 



A 



Yellow and blue .... 



J 27-85 observed. 

 ' * \ 29-1 calculated. 



1-25 



These experiments were not at all elaborate, and as their greatest 

 difference barely exceeds one per cent, of the reflecting-power of 

 white cardboard, they may be taken as furnishing a proof of the 

 correctness of the process employed. 



Thus far we have dealt with the combination of white (grey) 

 disks with those that are strongly coloured, and it remains to give 

 an example of the process as applied to two differently coloured 

 but not complementary disks. To test this matter it was necessary 

 to find two coloured disks having the same or nearly the same re- 

 flecting-power. In my collection I finally found two such disks, a 

 cyan-blue with a reflecting-power of 23-9 and a purple for which 

 ' the figure 23*3 had been obtained : these disks, when combined, 

 gave a scarcely perceptible flicker. Since then graded series of 

 yellow disks have been made, but it has been impossible to find 

 time to hunt up their equivalents in luminosity and make the 

 necessary determinations. 



This flickering process having answered so well, the procedure 

 was reversed and used with great advantage to facilitate the deter- 

 minations of the values of the grey disks executed in the ordinary 

 way ; in other words, the series of grey disks as made by myself is 

 not pure grey but has a slightly yellowish tint that makes estimation 

 of equality of luminosity a little more difficult than it ought to be. 



