August 1 8, 1881] 



NATURE 



369 



4919-8 have been affected, while 4923-2 has remained at rest. 

 That will give an idea of the way in which we really do find the 

 laboratory work and the observatory work, each coming to the 

 rescue of the other, each hel]nng us to understand something 

 which, without the other record, would be excessively difficult. 

 Tests supplied by the Absence of Lines from the Solar Spectrum 

 It is my conviction that many lines of the different chemical 

 substances are absent from the solar spectrum when that ab-ence 

 cannot be attributed to anything depending upon reduction in 

 the quantity of the substance present. In connection with this 

 point there is an experiment to which attention may now be 

 directed, because it is an attempt to imitate solar conditions 

 somewhat ; so that the inquiry i'i rendered possible as to whether 

 these lines may not owe their absence from the Fraunhofer Imes 

 to their being the product of a very low temperature, a temper- 

 ature which we cannot expect to find in the sun in any regions 

 where the pressure would be sufficient to enable any absorption 

 phenomena to take place. The point of the experiment is this : 

 There are bodies which we can render incandescent at low tem- 

 peratures. For iron, as we have already seen, we have to use 

 a coil, but such substances as magnesium, sodium, lithium and 

 the like can be volatilised at the temperature of the Bunsen 

 flame, and at that temperature we get a certain spectrum from 

 them. Now a great many different spectra have been recorded 

 by different observers for these bodies, and the question was, 

 could we get any independent method of determining which 

 lines were really due to high and which to low temperatures. 

 Now it is generally conceded that the temperature of the Bunsen 

 flame is lower than the temperature of an induction spark ; 

 and we have an arrangement by which we can pass a spark 

 between horizontal platinum poles through a flame in which 

 the subitance to be experimented on is volatilised. In this way 

 we can see what change in the spectrum is introduced by the 

 passage from the temperature of the flame to the temperature of 

 the spark. We can fill the flame with the vapour, say of sodium, 

 and observe its spectrum ; then when the flame is nicely charged 

 in the region between the two poles, we can pass a spark through 

 it, and by throw ing the image of the spark upon the slit of the 

 spectroscope we can first of all get a spectrum of the flame, and 

 then the spectrum also of that particular part of the flame 

 through which the spark is pa'^sing. Now we really have got a 

 good deal of light from that method of observation. In the 

 case of magnesium, for instance, the change is very striking 

 (see Fig. 42). 



The flame gives us a spectrum in which are seen two lines 

 corresponding with the two least refrangible members (i^jand b.^ 

 of a very prominent group of lines in the green part of the 

 solar spectrum, and associated with these is a less refrangible 

 line unrepresented in the sun, the whole forming a wide triplet. 

 On passing the spark this last line is very greatly enfeebled, if 

 not abolished altogether, for the very obvious reason that the 

 molecule which gives rise to it is dissociated more rapidly at the 

 temperature of the spark than it is at the temperature of the 

 flame, and as that line dies out another solar line (^'4) appears, 

 the three forming a triplet similar to, but narrower than, that 



Spark Spectrum 



Flame Spectrum 

 {End offtutings) 



Fig, 



42. — Flame and spark spectra of inagni 



produced in the flame alone. Kirchhoff showed that potassium 

 was not present in the sun, the line upon which he worked 

 being the red line which is seen when potassium is thrown into 

 a flame. The fact that we get that red line in the flame shows 

 that it is a line produced by a low temperature; the mole- 

 cule which produces the vibr;ition therefore may probably be 

 one which is produced at a low temperature. But when we 

 pass a spark through a vapour giving us that red line we do not 

 increase, but rather reduce, the intensity of the line, and we 

 bring a great many lines into prominence which were not s;en 

 before, and those lines, I believe we are justified in saying, do 

 exist among the Fraunhofer lines. In the same way we can 



colour the flame red with lithium, but the red line of lithium is 

 not in the sun ; but by passing a spark through lithium vapour 

 we can intensify the line in the yellow and the line in the blue ; 

 and the line in the blue is undoubtedly among the Fraunhofer 

 lines. Therefore it appears that we really can account for a great 

 many of these variations in the solar spectrum by simply assuming 

 that those lines which are absent represent molecular groupings 

 so complex that there is no part of the sun where their ab- 

 sorption could be visibly produced, cold enough to allow them to 

 exist. 



Test supplied by the Lines strengthened in Spots and Flames and 

 those seen in the Spectra of Two or more Substances 



It has already been pointed out that these lines, which have 

 been called basic lines, have been tested in two ^\'ays. In 

 the first place, a list of lines had been prepared from Angstrom's 

 tables and Thalen's tables, and then they had been discussed in 

 connection with the bright lines seen by Voung in his observa- 

 tions on Mount Sherman. The result was striking, inasmuch as 

 of the 345 lines which were included by Y'oung, only a small 

 number of which were seen in spots and storms, 15 of the lines 

 which were recorded as common to two substances by Thalen, 

 had been seen almost without exception, the only exception being 

 in the case of some of the spots. The attack was then varied 

 by taking 100 observations of sun-spots at Kensington, deter- 

 mining, without any reference to the basic niture of the lines at 

 all, the 12 most widened lines in each spot which it was possible 

 to observe ; and then taking, side by side with these observations 

 of the spots, too observations of flames from the rich store which 

 has been recorded by Prof. Tacchini of Palermo. Then again, 

 without reference to the basic character of the lines, to plot the 

 lilies down in each flame day by day. 



As a reminder we may again refer to the diagram already 

 given (Fig. 36). It will be remembered that the result was a very 

 remarkable one. We found the lines of iron (we limited ourselves 

 to iron) seen in the spots «ere few in number ; that the lines of 

 iron seen in the flames were still fewer in number, and moreover 

 that the lines seen in the flames were not the lines seen in the 

 spots. That was a result which might have been considered as 

 very extraordinary if we h:;d brought to it no other considera- 

 tions than those with which we were conversant ten years ago 

 when the work began. 



What we have to do now, then, is to find what has been the 

 result of this inquiry with regard to the basic nature of these 

 lines. Have we, as a matter of fact, or have we not, in these 

 most widened lines in spots, and the most brightened lines in 

 flames, picked out those lines which are common to two sub- 

 stances. The facts are these : — We have, in the first horizon of 

 the lower part of the accompanying map, the lines recorded by 

 Angstrom in his first memoir as common to two substances ; the 

 names of tlie tw-o substances being given below. In the 

 fourth h riion we have the observations of Thalen made a few 

 years after the observations of Angstrom. And in passing from 

 Angstriim to Thalen we pass from the temperature of the arc to 

 the temperature of the induction coil. Now it will be seen 

 that Thalen also gives us lines in some cases agreeing with 

 Angstrom's, in other cases extending the information given by 

 him, and in order to make this work as complete as possible we 

 have gone over this region with the arc as Angstrom did, and 

 with the induction coil as Tlnlen did, only we have had the 

 advantage probably of using a more powerful coil. In fact we 

 have used two coils — one so arranged as to give us the maximum 

 effect of tension, and the other the maximum effect of quantity. 

 In the first pl.ice it will be seen there is a general agreement 

 between the observations — an agreement marred only in appear- 

 ance here and there by the fact that in some cases the lines are 

 so near the position of air lines that it has been impossible to 

 make the observation absolutely complete. In other cases the 

 appearance of imperfection arises from the fact that lines which 

 are not seen at the temperature of the arc b.gin to make their 

 appearance at the temperature of the coil ; so that in a case like 

 that at wavelength 5017, for instance, where Angstrom gives 

 no line as common to two substances, yet Thalen does. We 

 find that both are right; that at the temperature of the electric 

 arc that line does not appear in one substance or the other, while 

 at the temperature both of the quantity and the intensity induc- 

 tion coil the line is certainly there. In Fig. 36 A represents 

 Angstrom's work, T Thalen's, and L Q and L I my own work with 

 the quantity and intensity coil. 



