Pottox & Lronarp—On the Spectra of Barium, &c. 235 
PoTassIuM. 
Photographs were taken with metallic potassium in an atmo- 
sphere of hydrogen, and also with solutions of potassium chloride. 
Potassium is characterized by a very feeble spark-spectrum, 
only two lines showing with one minute’s exposure either with the 
metal or a saturated solution; and with a 1 per cent. solution 
they are scarcely visible. Itis remarkable that the flame-spectrum 
is very intense; apparently the temperature of the oxyhydrogen 
flame, or even the Bunsen, gives a far more brilliant spectrum 
than the condensed spark. This is, no doubt, owing to the greater 
quantity of vapour produced. 
Plate XIX.—Spectra 4 and 5 give the spectra of the metal and 
a saturated solution of the chloride respectively. With the metal 
the intensities of the lines AA 3447°5 and 3446°5 are not nearly so 
great as in the case of the other two lines, so we have marked 
them 8, instead of 10, to show the difference. 
Quantitative Spectrum of Potassium Chloride. 
Whee Palguerey 
Leah. Persistency. 
4047-4 | 10 
4044-3 | 10 
3447°5 | = =((8) 
3446-5 | (8) o 
Sopium. 
Sodium chloride was used for the solutions, and the metal was 
photographed in an atmosphere of hydrogen. Sodium gives a 
well-marked spectrum of three pairs of lines; but, with the ex- 
ception of the D lines, they are not very persistent; and, as in the 
case of potassium, the sodium lines do not show with the spark 
nearly so strongly as with the oxyhydrogen flame, or even the 
Bunsen burner. It is also very remarkable that the D lines do 
