162 
PROFESSOR W. N. HARTLEY ON 
too great a cooling power. A candle or gas flame owes its shape to the rapid 
ascension of heated combustible vapour and air, or air and gas mixed, and the 
maximum temperature is to be found near the tip of the flame. The cross section of 
the flame near its tip shoidd therefore be sufficiently large to completely envelop the 
support and substance upon it ; lienee it will be seen that to have a support as small 
as possible is a distinct practical advantage. For some time a difficulty presented 
itself in the study of flame spectra of solid substances at high temperatures owing 
to the necessity which arises for providing an infusible material suitable as a support 
for the substance to be tested, capable of withstanding the temperature of the oxy- 
hydrogen blow-pipe flame, and incapable of chemical action upon metallic oxides. I 
formerly used strips of iridium for the alkaline earths and their salts, but they are 
quite unsuitable for use with several substances. 
I propose to jjlace on record a most convenient method of observing spectra with 
the oxy-hydrogen flame, and to describe a considerable number of spectra which were 
photographed preparatory to undertaking the study of spectroscopic phenomena con¬ 
nected with the Bessemer “ blow ” and the manufacture of steel generally. 
The flame of hydrogen, proceeding from a large lead generator, is burnt with com¬ 
pressed oxygen in a small Bunsen blow-pipe, so fixed that the flame is vertical. 
The blow-pipe measures 3 inches in length and fths of an inch in external diameter. 
The substances to be examined are supported in the flame on small plates of kyanite 
about 2 inches in length, of an inch in thickness, and ^th of an inch in width. 
This mineral, which is found in large masses in C°. Donegal, contains 96 per cent, 
of aluminium silicate, a practically infusible material. It w^as analyzed in m}’- laiio- 
ratory some years ago, and owing to the intractable nature of the mineral, the 
analysis was made with some difficulty. 
It is exceedingly difficult to pulverize it, but it readily splits into laminae. 
TJic Instruments and Method of Photography Employed. 
The instrument used for the first series of experiments had but one quartz prism 
of 60°, composed of right and left handed halves, each of 30°. The photographic 
plates used were “ Ilford rapid ” and Edwards’ Isochromatic Plates, 
A number of experiments were made with various sensitizers, such as erythrosine 
used by Waterhouse and by Mallmann and Skolick, and cyanine, employed by 
V. Schumann. Their use proved advantageous in rendering gelatine emulsion plates 
sensitive to the yellow and red rays. 
It was found that diphenylamine blue, used in a similar manner as, and mixed with, 
cyanine, rendered gelatine-bromide plates rather more sensitive in the region between 
E and F of the solar spectrum. Schumann has found that emulsions made with 
5 parts of silver iodide, precipitated along with 95 parts of silver bromide, ai-e also 
sensitive in tins ])art of the spectrum. 
