J. N. Lockyer—New Method of Spectrum Observation. 809 
The flame spectrum with the dispersion employed showed 
no Li line except the red one (A 67052). On passing the spark 
from the Holtz machine, the yellow line (A 6102-0) and the 
blue line (4 4602-7) appeared as bright as the red line. The 
same results were obtained on repeating the experiment with 
the large induction-coil. 
Kaperiment No. V1l.—Potassium nitrate was tried by the 
method previously described in Experiment No. 1. 
e flame spectrum consisted as usual of the red lines (A 7697 
and 7663) and the blue line (A 4045), very faint. 
The effect of the spark was to bring out the yellow lines (A 
about 5800), those in the green (A about 5840), and the red 
double (A 6946 and 6918) out of the flutings visible in the red, 
the double at 7697 and 7663 not being affected. The experi- 
ment was repeated with the induction-coil, and the same obser- 
vations made, with the additional one that the spark also 
slightly intensified the blue line. 
Experiment No. VIII.—On repeating the experiment with 
metallic potassium, the same phenomena were more markedly 
observed, the lines about 4 5800, and other lines more refrangi- 
le, were visible as very faint objects in the flame; they were 
much strengthened, however, by the passage of the spark. 
Experiment No. 1X.—Some potassium was volatilized by the 
spark in front of the slit of the sun-spectroscope and compari- 
Son of the positions of the lines with the Fraunhofer lines 
made. It is believed that 45829-0, 58020, 5782% are all 
reversed in the solar spectrum. The less refrangible member 
of the red double (4 6946) was next compared, and was un- 
doubtedly absent from the sun. These observations, however, 
are rendered extremely difficult on account of the fluted appear - 
ance of the yellow lines, and must be repeated with a stronger 
sun and the electric arc. The spectroscope employed had three 
prisms, one of 60° and two of 45°. ; 
Experiment No. X.—The flame spectrum of magnesium was 
examined, a green triplet was observed which was at first sight 
taken for 6. Measurements of the lines. however, showed that 
the less refrangible member was less refrangible than 6, and 
had a wave-length 5209°8, and that the other two members 
were b' and 8 respectively. A fresh charge of magnesium was 
put into the flame and the spark passed; the original triplet 
Was now no longer visible, the line at 5209°8 having vanished, 
but 5* was now seen forming with 6’ and 2 a triplet of similar 
form on a smaller scale. : 
_ According to Thalén, there are three blue lines of magne- 
Slum at w.l. 4481°0, 45865, and 47035. These lines were 
looked for in the flame with and without the spark. Without 
the spark only one line was visible in this region; its position 
Am. Jour. Sct.—TairD — Vou. XIX, No. 112.—Aprir, 1880. 
