24 ‘MR SWAN ON THE CONSTITUTION OF FLAME. 
cide with first one and then the other side of the spectrum.” ‘“ But further, the 
yellow space of such a spirit-flame spectrum is crossed by a bright fixed line— 
Sir David Brewster’s monochromatic ray. It is a beautiful example of the prin- 
ciples just pointed out in this method of horizontal analysis, being of much greater 
width than the rest of the spectrum, and recalling to the imagination the appear- 
ance of Saturn’s ring when nearly closed, and seen through a telescope of moderate 
power. This ray, from its superior breadth, must necessarily come from that 
pale, tawny light which invests the bright part of the flame.”* On this state- 
ment I will meantime only remark, that while, according to Professor Drapsr, 
the yellow space corresponding to the line D of Fraunnorer, has a much greater 
width than the rest of the spectrum, and while the inclination of the sides 
towards the red end of the spectrum is described as being “ quite rapid,” mere 
inspection sufficing to observe it, no such marked variation in breadth was observed 
in the more refracted portions of the spectrum. Professor DrarrEr says, he 
“satisfied” himself that even in the more refrangible regions there is the same 
want of parallelism by rotating the telescope on its vertical axis, so that the ver- 
tical wires in its eye-piece might coincide with first one and then the other side 
of the spectrum.” From this it is obvious that the supposed want of parallelism 
in the more refracted portion of the spectrum was so small as to be insensible to 
the unassisted eye, and required instrumental means for its detection. We are 
told that the breadth of the spectrum was ascertained by rotating the telescope on 
its vertical axis, so as to cause the wires in the eye-piece to coincide, alternately, 
with first one and then the other side of the spectrum; but we are nowhere in- 
formed in what manner the amount of the rotation was ascertained. It seems, how- 
ever, to be highly probable, that since it took place round a vertical axis, the readings 
in the two positions of the telescope would give the difference of azimuths of the 
two sides of the spectrum. It is to be regretted that Professor DRAPER has not 
described his method of observation more explicitly ; for I shall show in the sequel 
that the validity of his results depends entirely on whether the observations were, 
or were not, observations of azimuth. 
I will now describe my own experiments. The flame to be observed was 
placed behind a horizontal metallic slit, 0°05 inch in width, and of sufficient 
length to allow the entire breadth of the flame to be seen. To prevent the 
disturbing effects of currents of air, the flame was surrounded by a four-sided 
sheet-iron chimney, having on the side next the slit an aperture covered by a 
piece of fine sextant glass, through which the light passed to the slit without 
undergoing any irregular refraction capable of affecting sensibly the results. 
The prism used was kindly lent me by Professor Forpes. It is a fine flint- 
glass prism, the workmanship of Secretan of Paris. The instrument employed 
for observing the spectrum was an excellent theodolite, constructed expressly for 
* Tbid., pp. 106-7. 
