TJie Structure of Coloitred Bodies. 2 1 1 



map. Opposite such points in the right hand column of 

 figures we have the approximate thickness of the plate in 

 micromillimetres, and still further to the right the corres- 

 ponding undecomposed colour. On applying the map of 

 the spectrum of the breast feather of our humming-bird, it 

 is seen to correspond in colour composition with a line 

 crossing the spectral chart at a point indicated in the right 

 hand column by the number -485. This number, therefore, 

 represents the thickness of the plate in question in micro- 

 millimetres. This point, moreover, is opposite the orange 

 of the 2nd order, and we thus also ascertain the position of 

 the colour examined on the Newtonian colour-scale. 



From the foregoing example it is obvious that something 

 more has been ascertained regarding the structure of the 

 objects under observation than can be determined by the 

 microscope alone. The barbules, which to ordinary micro- 

 scopic investigation appear devoid of structure, are seen to 

 possess the property of colour-production. Examining 

 such colour phenomenon, under varying conditions and 

 with suitable appliances, it is seen to correspond in all 

 respects with the interference colours produced by thin plates. 

 We are, therefore, justified in assuming the same structure 

 as the cause of the colour in the barbules, and, therefore, of 

 attributing to the colour-producing portion of the feathers a 

 structural arrangement composed of thin plates. It has 

 been pointed out how an approximate measurement of the 

 thickness of such plates is indicated by the position of the 

 spectrum on the chart. 



In the above example we have an instance of a structure 

 exhibiting a marked optical effect, namely, the production of 

 colour. From the nature and properties of such colour we 

 have inferred the nature of the structure producing such 

 effect. If all ultra-microscopical structures possessed the 

 property of colour-production, and if the structural cause of 

 all colour-production were known, the determination of the 



