524 M. H. Helmholtz on the Theory of Compound Colours. 



a parallelogram the sides of which are constant, but whose angles 

 can be altered by turning theprism round the axis of the telescope. 

 The known propositions of elementary geometry teach us that 

 the superficial contact of such a parallelogram is a maximum 

 when its angles are right angles, the area decreasing more and 

 more as the obliquity of the angles is increased. Now as the 

 same quantity of light illuminates a smaller surface more brightly 

 than a larger one, the apparent brightness of the spectral image 

 must be a minimum when the image is a rectangle, that is, when 

 the refracting edge is parallel to the slit, and the brightness 

 must increase as the angle enclosed by the slit and the refracting 

 edge increases. 



The two legs of our angular slit, when looked at through a 

 vertical prism, give two spectra equally bright, inasmuch as the 

 refracting edge is inclined towards each at an angle of -IS" ; when 

 however the pi'ism is turned round the axis of the telescope, one 

 of the angles becomes greater and the other smaller, the relative 

 brightness of the spectra being thus caused to vary in any re- 

 quired degree. 



The brighter a spectrum is made in this way, the more closely 

 are its coloured bands pressed together : lest this should too 

 much prejudice the spectrum's purity, it is advisable to obtain 

 great differences of brightness in another way than that just in- 

 dicated. This is accomplished with great facility by placing 

 pieces of paper, oiled or not oiled, of greater or less thickness, 

 behind one of the slits. These permit only a small portion of 

 the incident light to shine through, while through the other slit 

 passes the unenfeebled light of heaven. 



When a field is obtained in the manner described, covered 

 with the mixtures of every two pure colours of the spectrum, the 

 observer readily convinces himself that the hues, particularly 

 those in the whiter portions of the field, cannot be estimated 

 while saturated colours stand beside them. 



It is therefore absolutely necessary to separate the portions 

 regarding whose colour we wish to form a judgement, from the 

 remaining ones. When the telescope is used in the observations, 

 the means of effecting this is very simple. Let the cross wires 

 be fixed upon the place in question, and let the observer recede 

 to a distance of one or two feet from the eye-end of the instru- 

 ment. From this distance only a small portion of the spectrum 

 is visible through the eye-glass, the colour of which may be 

 estimated apart from the disturbing influence of the dazzling 

 colours adjacent. If the observer be long-sighted enough, the 

 intersection of the cross wires is seen from this distance with the 

 naked eye, at all events l)y aid of a weak concave glass suited to 

 the eye. In order to rediscover the observed combinations with 



