Ma7\ 26, 1874J 



NATURE 



413 



for, as Sir Wm. Thomson has calculated, the atoms in a drop of 

 water are so small that if the drop of water were magnified to 

 the size of the eartli, the atoms would then be seen not larger 

 than cricket-balls or not smaller than shot. 



It must be clearly understood that I here refer to the true 

 atom and not to the atom of the chemists, the weight of which 

 they give as the "atomic weight.'' It may probably turn 

 out that this is often a molecule, sometimes a complicated one, 

 which grtat heat or electricity can divide, the latter some- 

 times more than once. It is clear that if this be so, then the 

 vapour densities as referred to the atomic weight will be " ano- 

 malous," because the true atom and not the chemist's atom is 

 in question at these high temperatures. 



It is now time for us to pass to the action of the spectroscope. 

 The spectroscope, as you know, is the instrument which enables 

 us to deal with either the refraction or the dilTraction of light ; 

 that is to say, by means of refraction or diffraction we sort out 

 the rays of any beam which we may choose to use into a spec- 

 trum, and wc then study by means of that spectrum the nature 

 and conditions of the substance which gave us the liglit. 



And there is more than this. Not only can we deal with the 

 giving out of light as light is being given out by this lamp, 

 or that flame, or that gas before me, but we can equally use the 

 absorption of light by various substances, thus studying the na- 

 ture and conditions of these substances. You know very well 

 that if this lamp, instead of having a shade of ground glass had 

 a red one, the light that would reach your eye would be red. That 

 simply results from the fact that the red glass stops in the main 

 all light but the red, and allows the red to reach your eye. That 

 then is a case of absorption, as the giving out of light by the wick 

 of the lamp is a case of radiation. 



What, then, does the spectroscope tell us with regard to the 

 physical differences in matter? It tells us that if we have 

 matter in a sclid state, that is matter the molecules of which 

 are large and are near together, agitated by the w-aves of 

 heat, or by electricity, we get a spectrum from it of a par- 

 ticular kind, called a "continuous spectrum," because the 

 spectrum is absolutely continuous, the red, yellow, orange, green, 

 blue, violet, are all there, as you see them in the rainbow ; 

 whereas, if we deal with a gas or vapour not too dense, that is 

 with a subst.nnce the atoms or molecules of which are smaller and 

 further apart than in the former case, similarly agitated by 

 electricity or, in some cases, by heat, you find that instead of 

 having what is called a continuous spectrum, you have a spec- 

 trum in which the light is not continuous, but broken. The 

 result of this broken condition is that we have light as it were 

 only here and there in the spectrum. We have in fact bright 

 lines representing a few images of the slit, instead of a rainbow 

 band, complete from the red to the violet, representnig continuous 

 images of the slit. This you see at once enables the spectroscope 

 to tell us the difference between the rare and the dense states 

 of matter quite independently of what that matter may be, and 

 whether we use radiation or absorption as the test ; since a sub- 

 stance with a certain molecular arrangement absorbs precisely 

 the same undulations as it gives out with the same molecular 

 arrangement. No matter what it is, the spectroscope at once 

 tells us whether this matter is in a gaseous or vaporous state, 

 in which case we have lines or bands ; or in a state in which the 

 molecules are nearer together, when we get a more or less com- 

 plete continuous spectrum. This at once partly explains why the 

 almost invisible long waves of the oxyhydrogen flame soon fill a 

 mass of the most refractory metal with waves of all lengths, until 

 it shines out almost like the sun. It would appear that mole- 

 cules or atoms, when once set vibrating by either long or short 

 waves, perform all the vibrations proper to them under the con- 

 ditions present . 



How then about the chemical differences? Here the infor- 

 mation afforded by tlie spectroscope is of a much closer 

 character. In the first place it tells us that if you take 

 any substance whatever in a state of gas or vapour, not 

 only do you get bright lines, which tell you that you are 

 dealing with a gas, but you get different bright Imes for 

 every substance, so that you not only know that ycu are 

 dealing with a gas cr vapour, but you know at the same 

 time what particular gas or what particular vapour. This is 

 qualitative spectrum analysis, as the effects depend upon the 

 quality of the atoms or molecules present. Further, we see a 

 change in the spectrum from simplicity to complexity, by which 

 I mean that the lines increase in number and broaden, and that 

 the bands become more complete and their channelled structure. 



where it exists, comes out better, as we pass from a low to a high 

 pressure. This is quantitative spectrum analysis, the change 

 depends upon the quantity of the atoms or molecules present. 



Again, the spectroscope at once enables us in the main (and I 

 say in the main, because I have already referred to the border- 

 land between the metals and the metalloids) to differentiate quite 

 as sharply bet\\een metals and metalloids as it does between 

 solids and gases. 



A metallic spectrum is always a line spectrum when we 

 employ electricity to produce the vapour. Only certain metals 

 give us line spectra at low temperatures : these are mostly 

 monad metals which vaporise easily. 



A melalloidal spectrum is only a line-spectrum when we 

 employ electricity. Long heat-waves in their action upon the 

 molecules only produce bands and channelled spaces. Thus the 

 vapour of sulphur has three spectra, two to be obtained by heat, 

 the line spectrum only being obtained by electricity.. 



Nor is this all. As we can distinguish the spectrum of a 

 metal from the spectrum of a metalloid by the appearance of the 

 spectrum, so also does the spectroscope enable us to see a 

 difference between the spectrum of a compound molecule and an 

 elemental molecule. Let me explain what I mean : — If we are 

 dealing with a metallic elenient, we get a spectrum of a 

 particular kind so sharply defined that when any one has 

 once seen it, he always knows that an atom of a metal is 

 being dealt with. In the same way w-hen we are dealing with 

 metalloids, the spectrum is generally so entirely distinct from 

 the spectrum of a metal, that when you have once seen the 

 spectrum of a metalloid produced by the long heat-waves, you 

 will always be able to tell it again, there is no possibility of mis- 

 taking it for the spectiumof ametal. So far we h.avebeen dealing 

 with the elemental molecules, or perchance atoms of metals and 

 metalloids, but we can take a compound molecule. Let us take the 

 combination between metalloids and metals, such as some of the 

 salts of strontium — the chloride of strontium, iodide of strontium, 

 and so on : here we have compound molecules, that is, molecules 

 no longer built up of one substance, but of two ; and the long 

 heat-waves, although they can set them vibrating and therefore 

 make them radiate light, do not shake them asunder as high ten- 

 sion electricity does. 



We find that the spectroscope is perfectly competent to 

 separate such spectra from all others, so that when we have 

 oncej seen the spectrum* of, say, iodide of strontium, we shall 

 for ever afterwards know that such spectra are given by such 

 a compound molecule as iodide of strontium. The same remark 

 applies to the compound molecules in which oxygen enters as 

 one of the substances. Such spectra closely resemble the 

 spectra of the metalloids, but the bands are farther apart and lie 

 nearer the violet as a rule, so that it is not difficult to distin- 

 guisli them. 



Now when we have to do with a compound molecule, that is 

 to say, with an association of two molecules or atoms of two 

 different chemical substances, we shall at once see that this 

 question of vibr.itions instantly comes into play; for if the 

 function of vibration, whether we deal with large molecules 

 and long heat-waves, or small molecules and electricity, is to 

 render mure simple what in the first instance was compound, 

 then we ought to get spectroscopic differences. 



Let us again take the iodide of strontium ; the spectroscope is 

 perfectly capable of letting us see the exact effects, not of every 

 degree of temperature which we employ, but of any great 

 difi'erences of temperature. We can follow each increase of tem- 

 perature by observing the lines or bands which disappear, or which 

 begin to be visible, as the case may be, as the temperature is 

 increased. And s\milarly, if we have a mixture at a temperature 

 ot dissociation, and gradually bring the temperature down until 

 association takes place, then also the spectroscope is just as 

 competent to help us as it w-as before when we were dealing 

 with .an increasing temperature. We find that as the tempera- 

 ture decreases in the latter case, the peculiar compound spectrum 

 to which I have already referred gets more and more vi.sible at the 

 same time as the elemental spectrum gets less and less visible : the 

 order being one of strict law absolutely capable of prediction 

 the moment you know what are the elemental line.s, and the 

 lines of any particular compound which longest resists the action 

 of pressure. 



Now this is extremely important in its bearing upon the 

 celestial side, so to speak, of this inquiry, and therefore if you 

 will allow me I will still further enlarge what I have said about 

 this distinction between the metals and the metalloids. 



If I take sodium vapour at a very low temperature and at the 



