TJie Structure of Coloured Bodies. 205 



entered upon, and have alluded to the necessity of employing 

 some method other than mere microscopic examination for 

 carrying on such enquiry. According to their action on 

 light all structures may be divided into three distinct 

 classes : — 



{a) Structures, the physical nature of which is such as to 

 allow of a visible image being formed of them by reflection 

 or refraction of light, and these, since they are amenable to 

 ordinary microscopic examination, I have characterised 

 as microscopic structures. As examples of microscopic 

 structures may be cited all such as are sufficiently large or 

 coarse, and of suitable optical density, or colour, in relation 

 to their environment, to allow of the formation of a per- 

 ceptible image. Their name is legion. 



{b) Structures which, from their physical nature, are in- 

 capable of so acting on light as to admit of the formation 

 by reflection or refraction of a visible image (or replica) 

 of themselves, yet can so modify light as to produce some 

 optical phenomenon which is characteristic of the structure 

 producing it. To this group belong all bodies which appear 

 structureless by ordinary microscopic examination, and yet 

 give rise to some optical effects, as reflection, refraction, 

 absorption, polarisation, and various interference pheno- 

 mena. This is the class to which I have applied the term 

 ultra-viicroscopic, since the microscope is either not appli- 

 cable for their investigation, or, if employed, is merely 

 used as an aid to some other method of observation, or to 

 observe some other feature of the object than its ordinary 

 image. 



Since most colour-producing structures belong to this 

 group, it is that with which we are the most concerned in 

 the present inquiry. 



{c) Finally we have abundance of evidence of the exis- 

 tence of structures, the physical nature of which is such as 

 to render them invisible, and incapable of producing any of 



