506 



Mr. It. H. M. Bosanquet on the 



be the case when the cloud is homogeneous. (P. 435.) Tyndall 

 observed the track of a beam in air 18 feet long. The normal 

 polarization was the same throughout the whole length of the 

 beam. 



What is to be considered is really the nature of the emission 

 at any given angle with the axis of the beam. Instead, there- 

 fore, of regarding the eye as fixed and the point of the beam 

 observed as variable, it is convenient to suppose a fixed point 

 of observation in the beam, and to imagine the position of the 

 eye varied, so as to observe rays making different angles with 

 the axis of the beam. We thus obtain the distribution of pola- 

 rized light as emitted from a fixed point of the beam, and instead 

 of the neutral point of Tyndall we have a neutral angle. 



I will take the description of three of TyndalPs experiments 

 from p. 437 ('Radiant Heat'). The laboratory was filled with 

 smoke of some kind, and the electric beam sent through the 

 smoke. The neutral angle and position of maximum polari- 

 zation were observed, and the smoke gradually allowed to dis- 

 perse. The observation of the neutral angle was then repeated 

 at intervals. Thus the earlier observations in each experiment 

 correspond to the more advanced stage (coarser particles), and 

 the termination of each experiment to the less advanced stage 

 (finer particles). Stage I. does not seem to have been reached; 

 and this might have been expected, as the finest particles of 

 smoke are probably coarser than the attenuated vapours of the 

 exhausted tubes with which the phenomena of Stage I. were ob- 

 served. I must remark that in the account of these experiments 

 the actual character of the polarization (positive or negative) is 

 not stated. I have assigned this by the consideration that, in 

 the most attenuated condition of the smoke, the polarization of 

 light emitted normally is always positive. 



Angle of emission for initial maxi-"l 



mum ])olarization (negative) . J 



Initial neutral angle 



Successive values of the neutral 

 angle as the smoke cleared 

 away 



Exp. 1. 

 Incense. 



12 or 13 



66 

 63 

 54 

 49 

 43 

 33 



Exp. 2. 

 Gunpowder. 



63 



50 



47 

 42 

 38 



Exp. 3. 

 Resin. 



12 

 J No neutral 

 \ emission. 



90 

 Less than 90 



Final direction of emission for maximum polarization (positive) normal to beam. 



The third experiment shows the passage from Stnge III. (in 

 which the polarization is everywhere negative) into Stage II. 



