combined with one equivalent of chlorine, the number of lines which remain in the chloride is large-over 60 per cent., for instance, in the case of Lithium and 40 per cent. in that of Sodium. While, on the other hand, in the case of elements with greater atomic weights, combined with two equivalents of chlorine, we get a much smaller number of lines remaining-8 per cent., for instance, in the case of Barium, and 3 per cent. in the case of Lead.

These interesting results suggested the extension of the research into the examination of the long and short lines of the metal visible in various salts. This work will be fully dealt with in subsequent chapters.

As in the experiments already recorded the presence of the chemically combined metalloid had its effect upon the spectrum of the metal, it seemed desirable to see what mechanical mixture would do.

Experiments were therefore made with mechanical mixtures. How the results obtained in this inquiry may in the future help us to make spectrum analysis quantitative is stated in Chapter VIII.

It must not be forgotten that in the atmosphere of the sun we have the most stupendous case of mechanical mixture which can be conceived. It seemed desirable, therefore, that the lines reversed in the solar spectrum should be examined from this point of view.

I shall attempt to hsow in Chapter X. how our knowledge of the sun has been increased by this examination.

This book is the property of COOPER MEDICAL COLLEGE, SAN FRANCISCO, CAL.

and is not to be removed from the Library Roca by any person or under ang it whatever.

CHAPTER VI.

ON THE SPECTRA OF SALTS.

It was stated in Chapter V. that we get as a result of the experiments there detailed the following facts established:

(1) When a metallic vapour is subjected to admixture with another gas or vapour, or to reduced pressure, its spectrum becomes simplified by the abstraction of the shortest lines and by the thinning of many lines.

(2) When we use metals chemically combined with a metalloid-in other words, when we pass from a metal to one of its salts (I used chlorine)-only the longest lines of the metal remain in the spectrum of the chloride the number being large in the case of elements of low atomic weight and small in the case of elements of high atomic weight, and of twice the atom-fixing powder of hydrogen.

In the present chapter I propose to show what has been done in the way of following up these preliminary results, as a preparation for obtaining those referred to in the next chapter. This, therefore, is a parenthetical

one.

In the first place several series of salts were used in which the atomic weights varied :

Ist. In each series.

2nd. In the associated elements in each series.

With this view the spectra of the Fluorides, Chlorides, Bromides, and Iodides of Lead, Strontium, Barium, Magnesium, Sodium, and Lithium were observed.

The method given in Chapter II. was used. The spark was taken in air without a Leyden jar.

§ 1. Lead.

In the case of the lead salts it was found that the Fluoride gave the most and the Iodide the least complicated spectrum of the metal.*

* The details of the experiments were as follows:

Plumbic Fluoride, Pb F2.-The eleven longest lines of the following wave lengths, from Thalén's observations, 4167'5, 4246'0, 4386'5, 5163'0, 5372'0, 4523 5, 55460, 56070, 6040 ̊0, 6059′0, and 6452'0, were seen.

Of these it is to be remarked that 4246'0 and 4386 5 are only seen for a short time when the spark first begins to pass, 6452'0 is very faint, and 5523'5 very short; so that practically the spectrum contains but seven distinctly visible lines.

Plumbic Chloride, Pb Cl2.-On observing the spectrum of this salt, it is found to have been simplified in the following manner:

The lines left are 4167'5, 5163'0, 53720, 5523'5, 55460, 5607'0, 6040'0 6059'0, and 6452'0, nine in number; 5523'5 has become excessively short, and 6452 0 rather brighter than it was in the fluoride.

Plumbic Bromide, Pb Br2.-4167.5 still maintains its brilliancy undimmed, 51630, 53720, 5546′0, and 5607'0 remain; 5523'5 is just distinguishable as a dot on the pole, but 6040 0, 6059'0, and 6452'o are completely lost, the spectrum thus being reduced to five lines.

Plumbic Iodide, Pb 12.-4167'5 is little or not at all altered in appearance, 51630, 53720, 5607'0 yet remain; 5546'0 has become a dot, and faint indications of 5523 5, in the same state but much fainter, are visible.

Expressing these results in another way, by making the non-metallic element unity, we have :

The lines in the spectrum of lead, then, increase in length and number as we ascend the above series, ¿.e. as the percentage weight of lead increases, as is shown in the second table. The fact may then be simply stated, that in the case of the salts of lead the complexity of the spectrum of the metal increases as the atomic weight of the non-metallic element with which it is combined decreases.

§ 2. Barium and Strontium.

The next metal experimented on was barium, an element of much lower atomic weight than lead, and in this respect occupying a position not very much above the mean atomic weight of the elements. It was soon found that the facts observed with lead did not com

pletely hold with regard to barium, although they include the phenomena presented by the chloride, bromide, and iodide. Even with these salts, however, the phenomena, though the same in kind, differ somewhat in degree. For instance, the same number of metallic lines was observed in all these salts, and between the appearance of the chloride and the bromide spectrum there was no appreciable difference. In the case of the iodide, however, there was a sensible change in the direction expected from the behaviour of leadi.e. the spectrum became dimmer, that is, exhibited a tendency to die out.

When, however, baric fluoride was examined, a different state of things was observed. Instead of the spectrum becoming more complex it became simpler, exhibiting in fact only the four longest lines of barium with any degree of distinctness, and these showing but little brilliancy.

Strontium behaved in the same way as barium, the falling off of the lines in the fluoride being very marked.

It will be seen from the above experiments, and from the annexed map of the strontium salts observed in air taken as a type of this group, that the general statement made for lead does not apply to barium or strontium, the chlorides, bromides, and iodides. of which metals are pretty equally volatile, while their fluorides are apparently not readily volatile at any temperature which I could employ.