22 J. RHEINBERG ON THE ORIGIN OF CERTAIN COLOUR 



But without entering into the detail of the thing, which would 

 be tedious, it could be worked out that if a certain colour is 

 extinguished for the dioptric beam, then for the diffracted beams 

 each other colour would be successively extinguished ; and if that 

 were so, irrespective of the appearance and tints of the diffraction 

 spectra, it might be assumed that their total effect would be 

 to produce a colour just the complementary to that given by 

 the dioptric or central transmitted beam. For the whole of 

 the diffracted beams ordinarily produce Avhite light, and if we 

 subtract all the colours except the one already subtracted from 

 the dioptric beam, then just the complementary colour to that 

 must appear. I made the experiment by stopping out the dioptric 

 beam with a little black spot on a cover-glass placed over the 

 objective. And the result corresponded with theory : the pre- 

 viously blue diatoms appeared yellow, the yellow ones blue, the 

 green ones purple, etc. This change into the precisely com- 

 plementary colours is one of the prettiest and most remarkable 

 experiments one can make with these diatoms. One word more. 

 In doing this it is assumed that only the diffraction fan of the 

 first order is taken up by the objective. If parts of the diffraction 

 spectra of the second order are present, they must be cut off, 

 because they naturally alter the colours. 



Summing up the cause of the colour phenomena of A. Ralfsii, 

 we have now seen that they are due to an interference effect, 

 that they depend upon the form of the diatom as a whole, on 

 the thickness of the diatom, and on the relative density — i.e.-, the 

 refractive index of the silex and the medium in which the diatom 

 is embedded. 



It may be asked. Why should only A. Bcdfsii present these 

 phenomena, and not other diatoms as well ? The answer is 

 that besides J. Ralfsii and other, if not all, forms of the genus 

 Actinocyclus, a great many other diatoms do show the same 

 phenomena, but in a less degree. You have only to put an iris 

 diaphragm above a |-inch or a 1-inch objective, and look at a 

 miscellaneous set of diatoms, gradually closing the iris, and you 

 will be surprised how many of them change colour, mostly turning 

 to a sombre brown or chocolate or pale yellow hue. Several forms 

 of JVavicula, Plewosigma, and Grammatophora^ and many others, 

 act in this way. In fact, it is a quite general principle, and the 

 reason why it does not occur to a greater extent in these forms 



