40 
EDITORIAL OBSERVATIONS. 
to the old method, was obtained somewhat doubtfully after three hours’ 
hard work. [The lecturer proceeded to illustrate the phenomena describe 
by Bunsen, by the aid of the oxyhydrogen lime light, running through the 
description of the phenomena again.] But this application which I have 
mentioned is, however, by no means all that this method can accomplish ; it 
cannot only do that speedily and easily, which we find a difficulty with other 
methods to perform, but it can do for us that which the old method never 
could. We can detect by this method quantities of the substances the 
smallness of which is perfectly astounding. Choosing from a number of 
beautiful examples which Bunsen gives us in his paper, I would refer you 
to this diagram, on which I have marked some of the most important, to 
show the delicacy of the spectra analytical method. ‘Soda: 3005035 th part 
of a raillegramme, or Tea o ssa osth part of a grain.’ These numbers seem almost 
absurd. I can scarcely tell you to-night how they were determined, but, in 
order to show you how delicate the process is, if we burn a piece of sodium 
weighing, say the T gth of a grain on a piece of paper, we allow the vapour to 
diffuse itself throughout the room. Supposing the whole of it went up in 
vapour, you would have to see how much of that sodium was present in 
every cubic foot of the air in this room. Then you must know how much 
air each of these gas lamps requires to burn a certain quantity of gas in a 
certain given time—for instance, five minutes; then bring it down to the 
time required to take an observation, say two seconds, and in that two 
seconds you will see there must have been a minute fraction of the To’ooth part 
of a grain present in that volume of air. In this way, which is a sort of 
rough description of the method which Bunsen used, he arrived at this 
astounding fact, that the 3005005 th part of a millegramme, or the tsoto ooooth 
part of a grain of soda can be easily detected. Soda is always present in the 
air; all bodies exposed to the air show the yellow soda line. If a book be 
dusted near a flame, the soda re-action will be seen. If we leave a piece of 
platinum wire out in the air for an hour or two, it will no longer impart no 
colour to the flame as the platinum wire did before, but it will give us the 
soda flame. If I pass it through my finger you will see the same result,— 
the yellow flame indicative of the presence of soda, which the platinum wire 
has taken from my finger. 
Of lithia: Toobooth part of a millegramme, or gooLouth part a grain, can 
be easily detected. Lithia was formerly only known to occur in four 
minerals. It is now found by spectrum analysis to be one of the most 
widely distributed elements. It exists in almost all rocks, in three cubic 
inches of sea and river (Thames) water; in the ashes of most plants, in 
milk, human blood, and muscular tissue. An experiment which Bunsen is 
very fond of showing is that of holding the end of a cigar (which our Ger¬ 
man philosophers generally have at hand), in the flame. You cannot see 
the lithium flame produced by it with the naked eye, but it is quite plain 
when we look at it through a prism.” 
By the above extract it will be seen that the process is 
marked by extreme delicacy or capability of detecting 
almost inappreciable quantities of matter; Bunsen having 
demonstrated the existence of one of the new metals thus dis¬ 
covered by him, namely, caesium, when one part thereof was 
mixed with a hundred million parts of another substance. 
By it, also, we are enabled to obtain a knowledge of the 
composition of the earth’s crust with a degree of accuracy 
