96 HarrLeyv—An Investigation of the Connexion between 
The very sensitive character of the potassium spectrum explains why so large 
a number of minerals, and also of precipitates obtained in analytical processes, 
have been found to contain traces of potassium. Thus in 169 various ores and 
minerals not supposed to contain this element, potassium was found.* 
CALcIuM. 
Spectra were photographed from (a) solid salts of calcium ignited in the 
oxyhydrogen flame, from (2) calcium metal burning in the flame of a Mecke 
burner, from (c) particles of lime in the same flame, and from (d) the condensed 
spark passed between electrodes consisting of thin plates of metallic calcium, 
and (¢) the spark without condenser, taken from solutions. Eye observations 
were made of the spark without condenser, passing from solutions of known 
strength. The wave-lengths are approximations only, though obtained from 
actual measurements on the photographs. They closely approximate in many 
cases to the measurements of the lines made by Eder and Valenta in the spectrum 
of the spark, and of Kayser and Runge in the arc. It must be noted that the 
measurements of the flame spectra do not appear to coincide with those of calcium 
oxide recorded in ‘“ Flame Spectra at High Temperatures,” notwithstanding their 
haying a common origin. ‘T'o account for these differences it must be understood 
that the latter measurements were made from the edges of bands, and the wave- 
lengths might differ widely, owing to variations in the spectra caused by the width 
of the slit, the dispersion employed being too small, the duration of exposure too 
long, and the quantity of substance introduced into the flame being large. In the 
earliest experiments, the dispersion was that of only one-quartz prisms of 60°, but 
subsequently a spectrograph with four prisms was in constant use. It is certain 
that the exposure of the plates in the one-prism instrument was unnecessarily 
prolonged, both bands and lines being obscured by the great density of the 
photographs. 
Calem—Oayhydrogen Flame Spectrum. 
Calcium nitrate, prepared from an especially pure precipitated calcium 
carbonate, was dissolved in alcohol, and made of such a strength that 1 c.c. 
corresponded to 0:004 gr. of calcium. Isochromatic Seed plates were used. 
The following lines were visible, but less refrangible lines were not photo 
graphed, the plates being less sensitive to the red than to the green rays :— 
Quantity of Calcium. 
0-0028 gr. } ; 
0-0020 ox MA 5543, 5517, and 4226-99. 
0:0016 er. The lines visible were \A\ 5543 and 5517. 
00014 gr. No lines visible. 
* Hartley and Ramage, Journal of the Iron and Steel Institute, 1897. 
