581 
E1) IT O K1A L OUSE R VAT IONS. 
burning hydrogen gas, gives a feeble illumination with great 
heat, is the best for the purpose. Now, let us imagine the 
apparatus prepared, and the spectrum thrown into a box from 
which daylight is excluded. The professors above mentioned 
take less than a grain of chlorate of sodium, mix it with a 
small portion of milk-sugar, and burn it in a corner of their 
laboratory furthest from the lamp. An eye keeps watch 
on the inside of the box, and presently when the vapour 
has diffused itself, and meets the flame, a bright-yellow 
line is seen to cross the spectrum, and remain visible for 
some minutes. This is called the sodium line, for when¬ 
ever sodium is present in the atmosphere of the lamp-flame, 
and however combined with other substances, that particular 
line never fails to appear, even if the quantity of sodium be 
not more than one twenty-millionth of a grain. It is 
found that each metallic base and metallic earth has its own 
peculiar line or lines: lithium shows a faint-yellow and 
sharp-red line; potassium, a red and a violet; strontium, 
four lines—two red, one orange, one blue ; calcium, a green 
and an orange line; barium, more than a dozen lines, which 
may be described as one half green, the other orange. And 
as with the sodium, each substance is always recognisable by 
its own peculiar line or lines,howeverinfinitesimai the quantity 
tested. In fact, it would appear that even the ultimate 
atom could hardly escape this mode of analysis. 
While carrying on their experiments the professors 
noticed certain lines on the spectrum for which they could 
not account, produced by none of the substances enumerated 
above, but by some substance to them unknown. Con¬ 
sidering hereupon, they ventured to assert that there existed 
in nature a fourth alkaline metal hitherto undiscovered. 
To it the name of “ caesium,” and of “ spectrium,” has 
been given (referred to by us at p. 22). It was found in 
the mineral waters of Kneuzuach, the evaporation of 
twenty tons of which yielding, however, only 250 grains of 
the platinum salt of the metal. 
Thames water, taken at low and high tide at Westminster 
Bridge, samples from the Chelsea and Lambeth Water 
Companies, from the New Biver, from the wells above the 
