BLACKWELL. — DISPERSION IN ELECTRIC DOUBLE REFRACTION. 667 



The values of 



[ (,«€ — ,»<q)a "I 

 (j* t — it„) D J 



are given in Table VI, which was made 



from data published by Mace de Lepinay in 1892 on the values of 



TABLE VI. 



Dispersion Factor of Quartz. 



The first line is obtained by interpolation from determinations by Mace de 

 Lepinay, Jour, de Pliys., (3) 1, 31 (1892). 

 The second line is calculated from the first. 



Differences between different plates, thus far unexplained, are elimi- 

 nated by using equation (3) in which the ratio of the observations to one 



of their own number, — , is formed for each photograph, instead of 

 x D 



placing values of x\ from different plates in comparison, as would be 



the case in using equation (1). The observations have all been 



reduced to one strength of electric field, that obtained with 8990 volts, 



but the actual voltages were very nearly this, so that the corrections 



resting on the law of squares are small. In this way the values of x\ 



given in Table V have been treated by (3) and the resulting values of 



the double refraction, (fj e — f'J),\ cs )> U1 terms of (// t — /v)(Z), cs ) ns one 



hundred per cent, have been tabulated and plotted near the beginning 



of this paper (Table I, and Figure 1). 



Jefferson Physical Laboratory, 

 Harvard University. 



2" Jour, de Phys., (3) 1, 31 (1892). 



