394 Prof. D. B. Brace on the 



second set of observations. Notwithstanding these precau- 

 tions, the mirrors had to be resilvered and the oil gained 

 access to the nicols and the strip and attacked the cement. 

 This showed itself by a thin capillary line gradually encroach- 

 ing on the field of view of the nicol and the final loosening 

 of the sensitive-strip within its cell. This placed a definite 

 limit to the number of observations which I was able to make 

 and also to the extent of the paths within the trough, on 

 account of the time necessary to make adjustments with the 

 entire number of mirrors. 



In most of the observations, I used a small telescope of 

 some two or three diameters. 1 found that my eye was most 

 sensitive if removed from the telescope during a reversal of 

 the trough, thus relieving it from fatigue during the interval. 

 On account of local disturbances within the liquid, the trough 

 was turned very slowly, the time occupied in a reversal being 

 usually a minute or more. It was then noted whether the 

 beam of light followed exactly the same path as before and 

 adjustments made, if necessary. This could be determined 

 by its position with reference to a definite mark. The eye 

 was then placed at the telescope, and the field carefully ex- 

 amined for any change in the two halves at some selected 

 point in its centre, on the dividing or vanishing line. 

 Settings for a match were also made by means of the arm p 7 

 and the index read by an assistant. While the edges of the 

 field might show a slight contrast in the two halves, the 

 central portion was very uniform and the bounding line, 

 practically, a vanishing one, so that settings could be made 

 to a high degree of accuracy. The strip used was the finest 

 I have ever seen, and gave a dividing line far superior to 

 any half-shade plane-polarizing system I have had the oppor- 

 tunity of working with. A half-shade " Lippich,'"' placed in 

 the same position, was entirely inadequate to the sensibility 

 required. This was partly due to the fact that such a system, 

 to give a vanishing line, must have in general a much larger 

 source than a point radiant. It might also be mentioned 

 that, if the half-shade system had been next to the polarizer, 

 instead, definition would have been impossible through such 

 a length of liquid. The use of such a thin strip, e. g. 0"1 mm., 

 in a liquid of the same index, makes it possible to invert the 

 order with respect to the active substance and still obtain a 

 vanishing line as well as perfect definition ; just as was done 

 in my previous experiment with a half-shade elliptical pola- 

 rizing strip placed after the substance examined. In the 

 preliminary observations made during April, light was sent 



