TRANSACTIONS OF THE SECTIONS. 5 
other and became obvious to sense; but on turning the plate to the 
red end, the bands separated and obliterated each other. Thus the 
phenomenon became a simple consequence of the undulatory theory. 
On Professor Powell’s Measures of the Indices of Refraction for the 
lines G and H in the Spectrum. By Sir Davip BREwsTER. 
After noticing the discussion between Professor Powell and himself 
on this subject, the author drew attention to another demonstration of 
the accuracy of his former statements. In the diagram No. I. he had 
drawn the group of lines round G, which is a single line well marked in 
the spectrum, and distinctly delineated in Fraunhofer’s map. ‘The real 
line G is not situated in the middle of the group, but much nearer its 
least refrangible side, and hence the index of refraction, as taken by 
Prof. Powell from the middle of the group, must have a greater index 
of refraction than the real line G, whose wave length in the interference 
spectrum had been determined by Fraunhofer. In reference to the line H 
the error is much more serious, as appeared from the diagram No. IL., 
in which Sir D. Brewster had represented the two remarkable bands in 
the violet rays, the least refrangible of which is distinctly marked in 
Fraunhofer’s map with the letter H. Between the central lines of 
these two bands there are no fewer than fourteen lines in the same 
map. All the observations made by Fraunhofer, both on the camera 
spectrum and the interference spectrum, apply to the central line H of 
the least refrangible of the two bands; but the observations of Prof. 
Powell, to which the author referred at the Newcastle Meeting, were 
all taken from an imaginary line bisecting the interval between these 
two bands, and therefore were of no value as physical data for testing 
the new theory of dispersion. Prof. Powell has recently given new 
measures for the real lines G and H, thus admitting the accuracy of 
Sir D. Brewster's former observations. 
On the Decomposition of Glass. By Sir Davip BREwsTER. 
The author has had occasion to examine the phenomena of de- 
composed glass, both of that which is found in Italy, of which he 
has received the finest specimens from Dr. Buckland and the Mar- 
quis of Northampton, and of other specimens recently found in 
making excavations among the ruins of the chapter-house of the 
cathedral of St. Andrew’s. In decomposed glass the decomposition 
commences in points, and extends itself either in planes so as to 
form thin films, or in concentric coats so as to form concentric films. 
When the centres of decomposition are near each other, the concen- 
trie films or strata which they form interfere with each other, or ra- 
ther unite, and the effect of this is that the glass is decomposed in films 
of considerable irregularity, their surfaces having a finely mammillated 
appearance, convex on one side and concave on the other. The films 
