414 



NA TURE 



[Mar. 26, 1874 



highest temperature that I can get on the earth either by the long 

 heat-waves or by means of electricity, I find that there is abso- 

 lutely no difference whatever in the molecular arrangement of 

 that sodium vapour at the extreme points. Spectroscopically it 

 is absolutely the same. 



Then if I take, not sodium but another element, such as iron, 

 I find it excessively difficult, by means of the heat-waves, to 

 shake asunder the molecules of iron and tlie diatomic or poly- 

 atomic molecules of iron vapour at all. But we know that by 

 electricity non-atomic iron vapour can be got ; and then we may 

 say, at all events so far as the lines in the spectrum are concerned 

 (I do not mean their position, but their general nature), that we 

 get a spectrum from the vapour of iron, similar in character to 

 that of the vapour of sodium ; but the spectrum has become more 

 complicated as we pass from the monad metal to one with a 

 higher atomicity. 



Suppose that, instead of taking a monad metal like a sodium, 

 with its few-lined spectrum, or a metal like iron, with its high 

 atomicity and its many-lined spectrum, we take a metalloid ; then 

 we find that those conditions no longer hold good. It is not too 

 much to say that in the case of the metalloids every change of 

 even low temperature brings about a change in the spectrum. 

 It is perfectly true, as I have said before, that by means of 

 electricity we can get a line-spectrum from most of the metal- 

 loids. But from tlie ordinary temperature to the electric 

 spark in the case of a metalloid, instead of getting the perfect 

 similarity that we did in the case of sodium vapour, we get an 

 equally perfect and equally beautiful dissimilarity ; so that whilst 

 we say that in the case of sodium we only know of but one 

 spectrum, in the case of sulphur, to take one case, we certainly 

 know of four. 



You must let me again remind you that when we era- 

 ploy electricity the spectra of the metalloids present exactly the 

 same appearance as the spectra of the metallic elements, such as 

 iron and sodium, and that it is only when we employ heat-waves 

 that those other changes to which I have referred take place. 



One word more, too, on the fundamental difference between 

 the spectrum of a metalloid and the spectrum of a metal on the 

 one hand, and the spectrum of a compound on the other. The 

 metalloid has a spectrum of channelled spaces or bands, some- 

 times to be found in the centr.al part, that is to say, in the green 

 part, or thereabouts, of the spectrum, whereas in the case of the 

 vapour of metals such as iron, and so on, we get bright lines 

 only, not bands ; and these lines increase in number generally 

 toward the violet, while in the case of the compound molecules, 

 such as iodide of strontium, to which I referred, we get a some- 

 thing which is half channelled spaces and bands, and half lines, 

 but in all the cases I have examined, excluding oxides, they are 

 limited to the red end of the spectrum. 



Let me attempt brietly to summarise what I have stated. With 

 electricity in the case of all elements we obtain line spectra ; as 

 we are here dealing with the most complete simplification of 

 matter that we can attain, let us call this the atomic spectrum. 



With heat we can obtain a continuous spectrum, from solids, 

 liquids, and some vapours ; with electricity we can even obtain 

 a similar spectrum from dense gases. Let us call this the 7nole- 

 cular spectrum. 



In the case of many of the metalloids we get, between these 

 extremes, a channelled space spectrum. Let us terra this the 

 sub-atomic spectrum. 



In the case of some compound molecules, we get by heat in 

 some cases, and by electricity in others, a spectrum which is dis- 

 similar from all these. Let us call this the compound atomic 

 spectrum. J. Norman Lockyer 



(To be continued) 



SOCIETIES AND ACADEMIES 

 London 



Royal Society, March .19. — Preliminary Notice of Experi. 

 ments concerning the Chemical Constitution of Saline Solutions, 

 by Walter Noel Hartley, F. C. S., Demonstrator of Chemistry, 

 King's College, London. 



The author has been engaged in investigating the above sub- 

 ject during the last eighteen months, and his experiments being 

 still in progress, he thinks it desirable to place the following ob- 

 servations on record : — 



In the examination of the absorption-spectra, as seen in wedge- 



shaped cells, of the principal salts of cerium, cobalt, copper, 

 chromium, didymium, nickel, palladium, and uranium, to the 

 number of sixty different solutions, it was noticed that the tinctorial 

 properties of the substances could be ascertained by noticing the 

 absorption-curves and bands, so that, provided water be without 

 chemical action, it could be foreseen what change would occur 

 on dilution of a saturated solution. 



The Effect of Heat on Absorption-spectra 



When saturated solutions of coloured salts are heated to 100° 

 C. there are (i) few cases in which no change is noticed ; {2) 

 generally the amount of light transmitted is diminished to a 

 small extent by some of the more refrangible (the less refrangible), 

 or both kinds of rays being obstructed ; (3) there is frequently a 

 complete difference in the nature of the transmitted light. An- 

 hydrous salts not decomposed, hydrated compounds not dehy- 

 drated at 100° C. , and salts which do not change colour on 

 dehydration, give little or no alteration in their spectra when 

 heated. 



Solutions of hydrated salts, and most notably those of haloid 

 compounds, do change ; and the alteration is, if not identical, 

 similar to that produced by dehydration and the action of dehy- 

 drating liquids, such as alcohol, acids, and glycerine, on the salts 

 in crystals or solution. 



A particular instance of the action of heat on an aqueous solu- 

 tion is that of cobalt chloride, which gives a different series of 

 dark bands in the red part of the spectrum at different tempera- 

 tures, ranging between 23° C. and 73° C. Band after band of 

 shadow intercepts the red rays as the temperature rises, till 

 finally nothing but the blue are transmitted. Drawings of six 

 different spectra of this remarkable nature have been made. 

 The changes are most marked between 33° and 53°, when the 

 temperature may be told almost to a degree by noting the ap- 

 pearance of the spectrum. Though to the unaided eye cobalt 

 bromide appears to undergo the same change, yet, as seen with 

 the spectroscope, it is not of so curious a character, the bands 

 being; not so numerous. 



With cobalt iodide a band of red light is transmitted at low 

 temperatures ; this moves towards the opposite end of the spec- 

 trum with rise of temperature until it is transferred to such a 

 position that it consists of green rays only. In this instance the 

 change to the eye is more striking when seen without the spec- 

 troscope, because the mixtures of red, yellow, and green rays, 

 which are formed during the transition, give rise to very beauti- 

 ful shades of brown and olive green. Thus a saturated solu- 

 tion at 16° C. wa3 of a brown colour, at - 10° C. it became of a 

 fiery red and crystals separated, at -1- 10° reddish brown, at 20° 

 the same, at 35° Vandyke brown, 45" a cold brown tint with a 

 tinge of yellowish green, at 55° ^ decidedly yellowish green in 

 thm layers and yellow brown in thick, 65° greenish brown, thin 

 layers green, 75° olive-green. An examination cf i m cobalt 

 salt has shown that there are two distinct crystahiae hydrates ; 

 the one formed at high temperatures has the formula 

 C0CIJ.2H2O, and is of a dark green colour; the other, which 

 contains a much larger proportion of crystalline water, is pro- 

 duced at a low temperature, and its colour is generally brown, 

 in cold weather inclining to red. 



The action of heat on solutions of didymium is characterised 

 by a broadening of the black lines seen in the spectrum, more 

 especially of the iinport,int band in the yellow ; and in the case 

 of potassio-didymium nitrate, this is accompanied by the forma- 

 tion of a new line. In the case of didymium acetate, which 

 decomposes with separation of a basic salt, the lines thickened 

 on heating. 



Thermo-clumical experiments 



Regnault (Institut, 1864; "Jahre»bericht," 1864, p. 99) has 

 shown that on diluting a saturated solution of a salt, as a rule 

 there is an absorption of heat, but in one or two cases he noticed 

 that heat was evolved. The change in colour that takes place 

 on the dilution of saturated solutions of cobalt iodide, cupric 

 chloride, bromide and acetate is very remarkable. There is 

 every likelihood that this phenomenon is due in each case to 

 the formation of a liquid hydrate. It is impossible of belief 

 that accompanying such a circumstance there should be 

 no measurable development of heat; and the author's experi- 

 ments have proved that in the above cases, at any rate, the 

 heat disengaged is very considerable, amounting, for instance, 

 on the part of cupric chloride, at least to 2,565 " units when I 

 gram molecule of the crystalline salt is displaced in its minimum 

 of water at 16° C. and brought into contact with sufficient tp 



