39 
numbers, it may be worth while to compare one of them, the 
yellow or the brightest part of the spectrum, with the value 
obtained by myself in a different manner. 
A prism cut from a thick plate of ice, with its refracting 
edge normal to the surfaces of the plate, was placed in a 
vessel having parallel ends of plate-glass filled with water ob- 
tained by melting a portion of the same plate of ice, to a depth 
of twice the thickness of the prism. The vessel containing 
the prism floating in water was placed so that a distant slit 
having a salted spirit flame behind it could be viewed with 
the telescope of a theodolite through the compound prism of 
ice, water and the plates of glass, and the image of the slit 
bisected, the prism was then depressed to the bottom of the 
vessel, and the slit as seen through the water and glass plates 
alone bisected. In this way the refraction out of ice into 
water was obtained unaffected by any small error in the de- 
termination of the angle of the ice prism, the impurity of the 
ice, and the want of perfect parallelism of the plates of glass. 
The resulting index for the line D out of ice into water at 
0° was 1:01952. According to Jamin (Comptes Rendus, Tom. 
XLUL p. 1191) if w,, mw, denote the indices of refraction of 
water at 0°C and ?°C respectively 
Lt, = pf — 0'000012573t — (0:000001929) #, 
up to 30°. With the aid of this formula, which yields results 
aereeing closely with those obtained by Dale and Gladstone 
(Phil. Trans. for 1858, p. 889), the indices of refraction of 
water observed at various different temperatures may be re- 
duced to their value at 0°. 
Of the observations by Fraunhofer, (Denkschriften der k. 
Akadenne zu Miinchen, B. V., 8. 223, 225), the third and fourth 
at 8° R and 91° R respectively, are the indices of a ray which 
he denotes by WV, coincident with the double line D of the solar 
spectrum. ‘The observations by Powell (The Undulatory Theory 
