J.— PSYCHOLOGY 225 



a third pile in which all the spots are grey. This part of the test is 

 supplementary to Part I, and undoubtedly does reveal any colour defect. 

 The colour blind presented the same heterogeneous collection of colours 

 as they normally do, the colour weak placed certain of the greys among 

 the greens, and so on. In this part of the test difficulty was sometimes 

 experienced by the normal colour group as to the correct pile into which 

 a card should go — -if a blue-green, for example, should be placed among 

 the greens or be in the left-over pile, if a reddish-brown should go with 

 the reds or be omitted. A further part of this test is to give one card 

 which has green and red spots on it of three different shades to the subject 

 and ask him to select out two cards in the pack which match these exactly. 

 This means selecting one card with red spots of the same shades, and one 

 card with green spots of the same shades. This proved difficult both to 

 normal and colour blind individuals. Some of the colour-blind gave 

 the task up as hopeless, only 2 out of those tested being successful. With 

 the N — group, 33 per cent, were successful, and with the N group 

 60 per cent, were successful. This part of the test, the sorting and the 

 matching, while of interest to an examiner, has not the same practical 

 value as Part I of the test, in which the break in the circle has to be 

 recognised. It could be dispensed with altogether, except in cases of 

 doubt, where confirmation of or further information about a defect was 

 required. 



The results from these two tests may be sufficient to give some idea 

 of the type of investigation which has been carried out. A similar 

 analysis of the results obtained with other pigment tests — the Stilling, the 

 Nagel, the Podesta, the Edridge- Green, etc., leads to similar conclusions, 

 namely, differences in the value of individual tests, inconsistencies in the 

 findings arrived at with any single series, and so on. As regards the 

 Stilling test, the complete investigation covers only the first three plates 

 of the seventeenth edition. These are found to give very satisfactory 

 results both in detecting colour blindness and in differentiating between 

 the two types. It is unfortunate that the formation of the figures used in 

 the plates is not clearer, because this gives rise to a number of trivial 

 mistakes which may be misleading to an inexperienced examiner. Such 

 errors are to mistake a 3 for an 8, or vice versa, or to confuse the figure 2 

 with 7. 



In the latest edition of the Stilling (the nineteenth edition) the same 

 figures are used, so that the same comusi ms appear. There are, however, 

 several new features, including the insertion of hidden numbers, which 

 are only visible to the colour blind. The results from this series are not 

 so clear-cut as in the earlier editions, and there is no reason to suppose 

 that it can yield more reliable or more consistent results than the others. 

 It is interesting to note, however, that the examiner is encouraged to ask 

 subjects to trace the shape of the figures with a wooden rod, to point out 

 dots of a similar colour to the one indicated, both useful devices in ordinary 

 testing, as also in cases of dissimulation or where it is suspected that the 

 subject has been coached in the test. Also, it is suggested that if in 

 doubt, the subject can be asked to read some of the plates through 

 coloured filters. This is a device which could well be followed with 



