486 



Professor Silvanus P. Thompson 



[May 17, 



First I will show you tlie transition tints of the first and second 

 orders in the bright field. That of the second order is much less 

 intense than that of the first ; and yet it is very sensitive, turning to 

 a green tint whilst the first order purple has only turned to a blue. 

 On the other hand, with reversed rotation of the analyser it turns to 

 red less rapidly than does the tint of the first order. 



Next I take the transition tints of orders I., II., and III. in the 

 dark field. These, though arranged, by means of superposed half- 

 disks of "quarter- wave" plates, to be optically equivalent to biquartzes 

 of two rotations, are really built up of selenite and mica. You will 

 notice how the tint of order I. surpasses in sensitiveness both 

 the others. I cannot here show you on the screen the means 

 by which I have compared the tint of order I. in the dark field with 

 that of order I. in the other set. Sufiice it to say that I find the tint 

 of order I. in the dark field — corresponding to 7 • 5 millimetres thick- 

 ness — more sensitive than that of order I. in the bright field, which 

 corresponds to 3 • 75 millimetres thickness. 



A method which was at one time supj^osed to be more precise, 

 was that of placing a spectroscope (or its prism) in front of the 

 analyser, and w^atching the motion along the spectrum of the inter- 

 ference bands which are then seen. My three pieces of crystal remain. 

 I introduce a slit in front of them, also a single film of quarter-wavo 



mica, and then a prism to 

 give the spectrum. This 

 prism (Fig. 9), by the way, 

 is a new sort of direct- 

 vision prism, having a 

 single very wide - angled 

 prism of Jena glass inclosed 

 in a cell with parallel ends 

 containing cinnamic ether 

 (first recommended by 

 Wernicke), a liquid which 

 has the same mean refrac- 

 tive power but widely diflfer- 

 ent dispersion. It is pre- 

 ferable to bisulphide of 

 carbon in several respects; 

 of cinnamon ; it is barely 



Fig. 9. 



Direct-vision prism iw j>rujection of spectrum. 



its odour is 



a delicate reminiscence 

 volatile ; and it is whiter than bisulphide. This prism, which is 

 shown also in plan in Fig. 10, was constructed for me by Messrs. R. 

 and J. Beck. It will be seen that the dark bands in the sj)ectrum 

 are nebulous and ill-defined. It was proposed to secure accuracy 

 by turning the analyser until they shift along to a definite point. 

 But their want of definition prevents precision. There is no advan- 

 tage in using the higher orders of tints which give more bands; for, 

 though the bands are certainly better defined, their progression across 

 the spectrum for a given amount of rotation is proportionally smaller. 



