2IO Dr. a. Hodgkinson on 



neglecting the effect produced by variation in the optical 

 density of the substance composing the plate, a given colour 

 is produced by a given thickness of plate. That is to say, 

 if the colour is known, the thickness of the plate can be cal- 

 culated. To ascertain the nature of the colour we employ 

 the only reliable test of colour composition, the prism. 

 Adapted to the microscope in the form of the so-called 

 microspectroscope, this shows the orange light reflected 

 from the feathers at normal incidence to have a composition 



■60 -515 



Spectrum of feather from breast of the ' ' Ruby and Topaz " Humming-bird 

 ( Chrysolampis mosqiiilus). 



indicated by the above spectrum. Such spectra are readily 

 mapped out on blank charts prepared for the purpose. 

 Since now, as remarked above, disregarding optical density, 

 a given spectrum is peculiar to a given thickness of plate, 

 it only becomes necessary to compare the obtained spectrum 

 with the spectra of thin plates of known thickness to learn 

 the thickness of plate-structure producing the spectrum 

 in question. 



To facilitate such comparisons I have constructed the 

 accompanying " Spectral Chart." (See coloured plate.) This, 

 as seen, allows of the immediate determination of all inter- 

 ference colours whether due to polarisation or produced by 

 thin plates, from the ist to the 7th order inclusive. We shall 

 subsequently refer to the construction of this chart. To use it, 

 it is merely necessary to slide the map of the spectrum of 

 the body under observation up the spectral chart, beginning 

 at the bottom, until on a level with a tranverse section of 

 the chart which shows the same colour composition as the 



