50 COLOUR IN NATURE chap, ii 



certain birds. It is incapable of existing in states 

 of oxidation and reduction, and contains copper and 

 not iron, so that it is unlikely that it is closely 

 related to haemoglobin. Carmine again is not 

 nearly related either to turacin or haemoglobin. 

 The spectrum test in this case seems to break down 

 utterly. In spite of numerous difficulties of this 

 kind, the literature contains numerous identifications 

 of pigments based only upon the use of the spectro- 

 scope ; thus a recent observer describes chlorophyll 

 in the skin of the horse ! In the following pages 

 identifications based only upon spectra have been 

 almost entirely omitted, for the evidence is con- 

 sidered insufficient. 



