August 22, 1895] 



NA TURE 



391 



ments he succeeded in getting coloured papers which are very 



good reproductions of the hues of the solar spectrum. These 

 pajieis have now been used for several years very extensively 

 in kindergarten and primary school work, and they are an im- 

 portant means toward the education of a new generation of 

 students to a true conception of colour, a more careful use 

 of colour terms, and a sharper discernment of colour percep- 

 tions. 



At a meeting of the Society of American Naturalists, held in 

 Boston, December 31, 1890, I read a jiaper in which was given 

 a more elaborate carrying out of the scheme which I had pre- 

 viously proposed. 



In order that any fixed scheme of colour nomenclature may be 

 of some practical value it must, of course, be readily understood 

 by people of only ordinary intelligence, and must be complete 

 enough to meet the ordinary wants of everyday life. There 

 must be something that is so completely fixed as to be perfectly 

 trustworthy for present and future needs. 



In the solar spectrum we have an invariable source from which 

 to derive our siiectrum standards, and upon these the whole 

 scheme is to be based. 



Since, however, the six spectrum standards do not give a very 

 extensive repertoire for common use, to say nothing of the needs 

 of the more artistic, it was proposed to introduce between each 

 two spectrum standards tw-o intermediate hues to be formed by 

 the union of the two spectrum standards in definite proportions. 

 Thus between orange and red would be introduced an orange- 

 red and a red orange. In the former red would predominate, 

 while in the latter orange would be more ]irominenl. Inasmuch 

 as these hues are only intended to be combinations of the 

 spectrum standards, it is not necessary, or even perhaj^s desirable, 

 that these hues be absolutely fixed. If, however, this is desirable 

 in any jiarticular case, it can be accomplished in a manner which 

 will be indicated subsequently. In addition to the two hues in- 

 troduced between each two standards it is also necessary to use 

 a violet-red and a red-violet (or two purples, a reddish puri)le 

 and a violet purple) to rejircscnt the actual combinations which 

 occur in nature. 



Ii is also very desirable that the standards I)e produced in 

 some material form in order that it be of any practical value. The 

 task of reproducing the brilliant hues of the solar spectrum in 

 pigmentary material or in glass is much more difficuli than (jne 

 not accjuainted with the matter would suspect. It would not be 

 difficult to select well-known pigments, and then determine the 

 wave-length which most nearly corresjionds to the hue of the 

 pigment ; but any number of such selections would not form a 

 symmetrical series of colour standards. The colours for such a 

 scheme being selected and their wave-length determined, the 

 other and more difficult problem is that of finding some com- 

 bination of pigment which will reproduce it. This task of 

 reproducing the spectrum hues was a very difficult one. It is 

 impossible to reproduce some of the spectrum colours with the 

 ordinary pigments either in hue or in (|uality. 



Almost at the very outset of this work in colour it was found 

 that it would be necessary to depend upon the somewhat fugitive 

 aniline colours for some of the standards as the only colour 

 material which would ajiproach the sjiectruni hues in brilliancy. 

 The difficulty of keeping the standards up to tone, so to say, 

 while using somewhat changeable material, is a serious incon- 

 venience but not an insuperable barrier. With the solar spectrum 

 recognised as the source to which we nnist always go to correct 

 our standards, the great difficulties of colour-study are met. 

 The most desirable thing now to be accomplished is the discovery 

 of some permanent colour material in which to reproduce the 

 spectrum standards. Some convenient form of tablet would 

 then be produced which could be supplied to all who are willing 

 to provide themselves with it, and to these all questions of colour 

 would be referred. The standards thus established, the inter- 

 mediate spectrum hues are determined by them. 



Now, by the use of the Maxwell discs in the standard colours 

 described above, we may fix upon definite proportions of each 

 which we will use for any other hue. If, for examjile, we desire 

 to introduce between red and orange two hues, we must first of 

 all know something of the relative effect of the two colours, and 

 combine them in inverse proportion to what we are accustomed 

 to call the value of the colours. The colour which has the lowest 

 value will require to be u^-ed in larger proportion than the other. 

 In this case we may take a red and an orange disc and put them 

 together in the manner above described. For convenience of 

 measurement, a disc just a little larger than the coloured discs, 



with the margin graduated into one hundred dejgrees, is placed 



behind the coloured discs, and the sectors adjusted as desired. 

 As the red has the lowest value, more of the red disc must be 

 exposed in order to produce an effect equal to that produced by 

 the orange. If it is desired, therefore, to introduce two hues 

 between red and orange, we must still more increase the pro- 

 portion of red in the combination which we wish to be most like 

 the red. For our orange-red we may use 70 per cent, of red and 

 30 per cent, of orange, and for our red-orange 59 per cent, of red 

 and 41 per cent, of orange. By making a scale of values for the six 

 standard hues, we may combine them in the manner we have just 

 illustrated and form two hues between each of the standards, and 

 two more by combining red and violet. These twelve hues, 

 with the six standards, give a sufficiently large variety of hues 

 for practical purposes. 



For purposes of colour education, however, it must be home 

 in mind that pure spectrum colours are not often seen either in 

 nature or art. And while it is very important that the student 

 should be taught the spectrum colours at the outset of his educa- 

 tion in order to establish some accurate knowledge, derived from 

 the only source of accuracy, the solar spectrum, it is also im- 

 portant that he should become familiar with the effect produced 

 by the mingling of these spectnmi hues with the light reflected 

 from other oljjects, as well as the effect of shadow upon the 

 colours themselves. The mingling of white light with any colour 

 produces a tint of that colour. The tints are what we most often 

 see in all exce])t the most brilliant colours of flowers, not generally 

 of the standards but the intermediate hues. On the other hand, 

 when a coloured object is seen in shadow, or, what is more com- 

 mon, when the coloured surface is so irregular as to reflect here 

 colour and there give no reflection, the effect is to produce a 

 shade of the colour. In foliage the prevalence of shades is the 

 rule, whether we consider the individual leaves or the masses of 

 foliage. -\ knowledge of these effects is best acquired by the 

 use of a very few tints and shades of each hue. Any convenient 

 number of tints and shades can of course be designated, but a 

 few will serve all the purposes of ordinary educational work. In 

 their educational papers the Milton Bradley Company use the six 

 spectrum standards, twelve intermediate hues, including the 

 combinations of red and violet, two tints and two shades of each 

 of the pure colours, thus giving in all a range of ninety different 

 modifications of colour. With these are used black and white, 

 together w ith a variety of greys. The facility with which young 

 children learn to distinguish and designate colour is really quite 

 surprising. 



But a still larger proportion of the colour effects of nature and 

 art than those produced from either tints or shades are the 

 result of both light and shadow combined with colour. This 

 effect has been well enough described by the term " broken 

 colour." In order to acquire familiarity with this eflect, it is 

 desirable to use a series of broken standards, if not also of the 

 twelve intermediate hues. These should be made, as should also 

 the tints and shades, by using proportions which take into 

 account the value of the colours, and, above all, the proportions 

 of white and black used should be such as to avoid destroying 

 the characteristic eftect of the colours. Each of these broken 

 standards may have its tints and shades like the standards them- 

 selves by increasing the amount of white or black which is 

 combined with the colour. 



But perhaps the most interesting jioint in connection with the 

 introduction of definite colour standards will be the possibility 

 of talking about colour in a definite language. Without such 

 standards this has been impossible. By the use of the Maxwell 

 discs made in the standard colours we may easily determine the 

 composition of any colour. This is a great convenience in the 

 description of colours, for it renders it possible when it is 

 necessary to give an exact meaning to any colour term. Only 

 with standards which can be accurately fixed is this possible. 

 The use of such terms as vermilion, emerald green, ultramarine, 

 chrome yellow, and similar terms as a basis of colour analysis is 

 exceedingly impracticable, since even these terms, although by 

 far the most definite terms in common use, are ipiite too variable 

 to give results which can be of any real value. For the sake 01 

 convenience, the first letter of each colour is used as the symbol 

 of the colour in all formulce in which the analysis or composition 

 of colour is expressed. N is used for black, to avoid the 

 repetition of B which is used for blue. 



The following formul.-v will illustrate the practical application 

 of the idea and the value of the sjiectrum standards in determin- 

 ing the composition of colours. They will also be of interest as 



NO. 1347, VOL. 



52J 



