| 
| 
. 
Fal ea 
Ser Re Mae ka kee Se Re 
eee ies eC ka ee ee 
C. 8. Hastings— Optical Constants of Glass. - 275 
question are as 9:6:5 nearly; hence if this relation holds 
approximately for all optical glasses, as is probable, an achro- 
combination good for one temperature is good for all 
others within moderate limits. 
Fizeau found, in the researches before alluded to, values for 
the refractive increment for 1°C., not unlike our own, namely : 
Crown glass (zinc), sp. gr., 2°626, Ap>=0°00. 
Common flint glass, sp. gr., 3°584, Ap>=2°6. 
Dense flint glass, sp. gr., 4°14, Ap=6°87. 
The magnitudes of the quantities £ show at once the importance of 
observing the temperature of the prism in every accurate deter-* 
mination of refractive indices, neglect in so doing generally viti- 
ating the fifth decimal place. It may be remarked, too, that 
ordinary variations in barometric pressure cannot be neglected 
when it is desired to limit the errors to the sixth decimal place. 
The prisms IV and V were not studied to determine the 
Temperature, 20° C. Barometer, 30 inches. 
Line. IV. Vv. 
A 1°603945 509607 
B 1607306 611584 
C 17609041 512580 
“Ds 1°613843 515288 
5614 1°616333 [516673 
1°620103 518719 
F 1°625751 521696 
G 1°636702 1°527300 
h 2 284 1°530075 
3951 1°647048 1°532381 
The probable errors were not computed, for the reason that the 
temperature corrections were assumed; but as regards their 
accuracy, the values under IV are perhaps quite as good as 
those which go before; those under V, on the other hand, 
have much larger errors, though probably in every case below 
two units in the fifth place. Ti is finds its explanation in the 
fact that the material was not homogeneous, thus giving less 
perfect definition than the others, though the surfaces of the 
ing the theory of the astronomical objective. 
Johns Hopkins University, January, 1878. 
