ESTIMATION OF CARBOHYDRATES 217 



angles to each other: i.e. the vibrations of each ray which are trans- 

 mitted are now only in one plane. 



By employing a rhombohedron of Iceland-spar, cutting it across 

 through its obtuse angles, polishing the cut surfaces, cementing 

 together these cut surfaces with Canada balsam, and blackening the 

 longer sides, a prism is obtained. On passing light through this prism, 

 the ordinary ray is totally reflected by the cut surfaces and absorbed 

 by the blackened side, whilst the extraordinary ray passes through and 

 emerges in a direction parallel to the source of the light. Such a 

 prism is termed, after its discoverer, a Nicol prism. 



In a polarimeter two Nicol prisms, mounted in line with one another, 

 are employed. The first is fixed, the second is capable of being 

 rotated. Light is passed through the first prism (the polariser) and 

 reaches the second prism (the analyser). If this second prism be ex- 

 actly parallel to the first, the beam of light will also pass through it ; 

 if it be not exactly parallel but inclined at an angle, less light will pass 

 through it ; if the second prism be at right angles to the first, or 

 crossed, the light is entirely cut off. 



By interposing between the prisms, set parallel to one another, a 

 solution of an optically active substance, the amount of the light is 

 diminished, but it can be brought to its original intensity by rotating 

 the analysing prism. The amount of rotation necessary to effect this 

 corresponds with the power of rotation of the solution. As the an- 

 alysing prism is mounted on a graduated circle, the number of degrees 

 rotated can be measured. This is the rotatory power of the solu- 

 tion. 



The determination of equal illumination of light in such an instru- 

 ment before and after its passage through an optically active solution 

 is very difficult and the readings are erroneous. Several devices have 

 been adopted to overcome this difficulty, the simplest being that of 

 Laurent. Laurent placed behind the polariser a quartz plate of special 

 thickness and of such a size that it covered half the field. This quartz 

 plate divides the ray of light passing through it into two rays, one of 

 which is retarded by half a wave length and therefore reversed in direc- 

 tion, whilst the other is unaffected. The resultant ray formed on 

 emergence by their fusion will be vibrating in a plane at an angle to the 

 original plane, i.e. the polarised light passing through the quartz plate 

 is rotated through a certain angle. Thus, if AO be the original plane 

 before passage through the quartz plate, it is resolved into AC and AD. 

 Supposing AD is retarded and reversed, then the components AC and 



