272 WRIGHT: MEASUREMENT OF REFRACTIVE INDEX 



Method 2. The thin plane parallel plates can also be used with a 

 small 60° glass prism (10 x 10 mm.) and the measurements made by 

 the minimum deviation method. In this case both a telescope and a 

 collimator are required. The liquid is held in place by capillarity 

 and prevented from running up along the ridge into the prism support 

 by a notch in the glass plate, as in figure 1. The glass plates are not, 

 however, silvered as in the first method. In this form the prism is 

 somewhat similar in form to the glass prism used first by H. G. Madan"* 

 for the measurement of the refractive indices of piperine. Madan 

 cemented two thin plane glass plates to the sides of a prism of dense 

 flint glass, allowing the glass plates to project above the top of the glass 

 prism. Into the angle between the glass plates he poured the melted 

 piperine and thus obtained the desired prism of piperine. The same 

 purpose was accomplished later by H. E. Merwin^ of this laboratory by 

 fusing together at one end two strips of cover glass or object glass under 

 an angle of about 60? Such strips are not perfectly flat and plane- 

 parallel and the degree of accuracy obtainable by their use is, of course, ■ 

 less than it would be, were plates of better quality used. The cover glass 

 prisms are, however, sufficiently accurate for ordinary purposes and can 

 be prepared in a few minutes. 



Method 3. By the use of a small accurately ground 60° glass prism, 

 5 mm. on a side and 10 mm. long, and three thin plane parallel glass 

 plates (5 X 20 X 2 mm.), a three sided liquid 60° prism can be obtained 

 and the measurements made by the very accurate method which 

 Gifford' employed so successfully with large glass prisms. 



In the last three methods the regulation of the temperature is not 

 easy but the temperature at which the measurements are made should 

 be carefully noted as a rise of 2° or 3°C. produces an average decrease 

 of 0.001 in the refractive index of a liquid. The refractive index of 

 liquids changes, moreover, on an average 0.001 in refractive index for 

 a change of 6 to 12 fx/j, change in wave length. Measurements, there- 

 fore, even to the fourth decimal place have little significance imless 

 the temperature be accurately recorded and monochromatic spectral 

 line sources be employed. 



Method 4. Measurements with the Abbe-Pulfrich crystal total-refrac- 

 tometer (accurate to one or two units in the fourth decimal place). 

 The methods for use with this refractometer are described at length 

 in the standard textbooks and need not be repeated here. Several 

 new details of manipulation, however, which have been found of value, 

 may be recorded briefly. The best readings are made when the light 

 enters at grazing incidence. In this case the field is divided into two 

 halves, the one light, the second dark; the boundary line between the 

 two is so sharp that accurate settings are easily made. With a single 

 drop of liquid, however, this method of grazing incidence has not been 

 applicable heretofore and recourse h&s been had to the method of total 



3 Jour, of the Chem. Soc. Transactions, 79: 922-927. 1901. 



* J. GeoL, 20: 495. 1912. 



' Proc. Roy. Soc, 70: 329-340. 1902. 



