June 4, 1874] 



NA TURE 



ATOMS AND MOLECULES SPECTRO- 



SCOPICALL Y CONSIDERED * 



II. 



1 now pass on to another part of my subject. 



7. Wlien Imo temperatiirts are employed it is generally achnuw- 

 ledged that there is an important difference m kind between the 

 spectra of metals and those of metalloids, taken as a luholcf 



SpectroscopiciUy it is more easy to define the difference be- 

 tween these two great cLisses of metals than the chemists among 

 you would imagine. I will ask you to take the spectrum of 

 the third class of stars as being as good a representation of the 

 spectnmi of a metalloid as anything I can place before you. It 



is rhythmic, the other two are not. It is a " channelled space " 

 spectrum.* That defines a metalloid spectrum ; and a similar 

 spectrum in the case of hydrogen is referred by Angstrom, 

 Stewart, Schuster, and others to an impurity. I have before 

 referred to temperature and told you that the temperature of a 

 Bunsen burner is enough to set an atom of sodium free from its 

 combination with chlorine and make its vapour give us a bright 

 line. I have told you we cannot do this in the case of iron and 

 other substances. We may say then that we have there a first 

 stage of temperature. Many monad metals give us their line spectra 

 at a low degree of heat. Take some dyad metals such as zinc 

 and cadmium ; this first st.age of temperature will only make them 

 red or white hot, a much higlier temperature is required to drive 

 them into vapour. We get the hne spectrum from sodium ; do 



Fig. 3. — Copy of a photograph of the long and short lines of iron between wave-lengths 4,000 and 4,, 



we get that from cadmium when we have melted cadmium ? 

 We do not. That is an excessively important point. The first 

 stage of temperature, which gives you a line spectrum in the case 

 of sodium, is powerless to give you such a spectrum in the case of 

 cadmium. 



A second stage of heat at least is therefore required to get 

 a line spectrum. If I take sulphur, dealing with it by means of 

 absorption, and heat it, I get a continuous spectrum at the first 

 stage. I increase the heat to the second stage, what do I get 

 then ? A line spectrum, as I do in the case of sodium ? 

 No ! A spectrum like that of the star in the constellation 

 Hercules, not a line spectrum at all. I apply still a higher, 

 a third, stage of temperature and then I get a line spectrum. 

 In the caseof the metalloids we have thus three stages of heat with 

 three spectra. If there is such a thing as a particle at all, are 

 we not justified in asking whether there is not some difference 

 between the "particular" arrangements of the metalloids, from 

 those of the metals ? and some connection between temperature 

 and the "atomic weights" of the chemist? 



Before I go further I will throw these results into a tabular 

 form, which will show you that through these various heat stages 

 in the case of metals like sodium there is a great preponderance 

 of line spectrum, and in the case of metalloids like sulphur there 

 is a great preponderance of channelled space spectrum. J 



Fifth stage — spark 

 Fourth stage — arc 

 Third stage — white 



heat 

 Second stage — bright 



red heat 



icontinuous ab- 

 sorption in the 



tine 



channelled space 

 channelled space 



channelled space 



continuous ab- 

 sorption in the 

 blue 



8. I next state that A compound particle —that is a particle con- 

 sisting of two distinct elements — has a zibration which is as peculiar 

 to itself as the vibration of a particle of an element is peculiar to 



* Continued from p. 7. 



t Since this lecture wa 

 ence of new spectra of sodiu 

 known ones of the metalloid 



I Since this lecture was del 

 much further, and it seems o 

 and chemists, as when it con: 

 spect 

 betw 



deli\ 



red Prof. Roscoe has established the exist- 

 nd potassium closely resembling the well- 



d I have carried this branch of the research 



ell deser\'ing of the attention of physicists 



I be acknowledged that different classes of 



do truly represent different "particular" aggregations, the contrast 



n the extremes of metals and metalloids will be beyond question. The 



ult marked thus U I have added from later work.— J N. L. 



itself. Thus the salts of strontium have each a distinct spectrum. 

 Take the particle of NoO^. The absorption spectrum of this gas 

 you now see on the screen. This particle has a vibration quite of 

 its own. Now it is a gas which it is perfectly easy to dissociate. It 

 is easy to turn it from ^X>^ to NOj. We introduce a new spec- 

 trum. These facts — and they might easily be multiplied— show 

 then that a compound particle is a perfectly distinct physical 

 thing, with vibrations, rotations, and free paths of its own. There 

 is no apparent connection between the vibrations of a compound 

 particle and those of any of the substances which make up that 

 compound particle. 



9 . I now come to another important point : On the whole certain 

 ki}ids of particles affect certain ) arts of the spectrum. Take the 

 bright lines of the metals ; if you were to mix together all the known 

 metals in the sun, make a compound which should consist of all of 

 them, put it into the lower pole of an electric lamp and photograph 

 the spectrum, then you would find the majority of the lines 

 would be in the violet end of the spectrum, scarcely any in the 

 red end. That is the re.tson why the spectrum of the sun, which 

 contains so many of the metals, is so complicated in the violet. 

 If you combine a metal and a metalloid, you will find, in many 

 cases at all events, that the vibrations will lie in the red end of 

 the spectrum ; you will also find that there is a connection between 

 the atomic weight of the metalloids and the region of the spec- 

 trum in which their lines appear under similar conditions. 



You have, in fact, simple particles and short waves, compound 

 particles and long waves. Nor is this all. In many cases we 

 find both ends of the spectrum, and in many cases the more re- 

 frangible end only, blocked by continuous absorption. This 

 occurs so often in absorption spectra that one is led to suspect 

 that it is due to some arrangement of particles. 



10. Here is another proposition : /// the case of metalloias, and 

 compound gases containing them, the spectrum to a large extent 

 depends upon the thickness of the vapour through which the light 

 passes, and often, if not invariably, the absorption increases to- 

 wards the red ind as the thickness is increased. 



Here is one of the points of the most extreme theoretical im- 

 portance, and one about which least is known. There is a state- 

 ment in Prof. Maxwell's book, that it you take a metallic vapour 

 and employ a great thickness of it, you will get from it the same 

 spectrum as you would from a small thickness of great density. 

 '1 his is Prof. Maxwell statement ; I venture to think that 

 this is somewhat doubtful, for in questions of thickness the 

 spectroscope can offer the physicist a million of miles or a 

 millimetre to work with, and one would think that such a 

 difference should be enough. If I had a tube with a bore 

 of the size of the lead in this pencil, and had some hydro- 



* By an oversight last week the ilIu;trjtion here referred ti 

 instead of the present one. — J. N. L. 



inserted 



