ipis.] MARSHALL-BANKS— NEW NEPHELOMETER. 179 



empirical work the elimination of this factor is not of very great 

 importance, the theoretical consideration of the problem is greatly 

 simplified thereby. 



As Wells states, the opalescence of a liquid containing a definite 

 amount of substance in suspension will, owing to the greater total 

 reflecting surface, increase with the continued subdivision of the 

 particles until these reach a limiting size. Rayleigh has pointed out 

 this fact in a mathematical dissertation on the blue color of the sky, 

 stating that as the particles approach the size of a wave length 

 of light their reflecting power decreases. He shows that for very 

 minute particles the amount of light reflected should vary as the 

 sixth power of their radius. The maximum opalescence of the 

 solutions as used in a nephelometer seems, however, to be devel- 

 oped when the particles are much smaller than a wave length of 

 light — in fact of ultramicroscopic size. 



The amount of reflected light lost through absorption is also a 

 function of the number and size of the particles. 



It is evident that as the refractive index of the medium ap- 

 proaches that of the particles, the amount of light reflected will 

 decrease until, when the two refractive indices become equal, there 

 will be no reflection. This phenomenon may be observed if pow- 

 dered glass be suspended in a mixture of carbon disulphide and 

 benzol. 



With a view to determining some of the underlying laws of 

 opalescent solutions, we undertook to design a nephelometer better 

 adapted both to theoretical and to practical work than those in use 

 at present. By using equal columns of suspension and actually 

 measuring the reflected lights with a suitable photometer, not only 

 is one of the variables eliminated, but also we are enabled to de- 

 termine the absolute ratio of the lights reflected by various sus- 

 pensions. The photometric part of the apparatus consists of a 

 wedge of neutral tinted glass by which the light from one of the 

 suspensions may be controlled; and a suitable optical arrangement 

 for observing the two beams of light. A Lummer-Brodhun prism 

 would serve this purpose admirably, but by a simple arrangement 

 of mirrors, a field far more sensitive than that of the Duboscq 

 colorimeter may be obtained. 



