TRANSACTIONS OF SECTION A. 453 



and concentric envelopes, and the spectra of these flames differ in different parts of 

 ?he arc Thus, in a photograph of the flame given by manganese the line at wave- 

 lena?h 4-^34 5 occurs without the triplet near 4030, while m another the tnplet is 



^"Xetes^wbtl'ai^'Stid most readily in the arc are generally those the 

 absorption of which is most developed in the flame ; thus the manganese triplet in 

 ?hevSet is reversed in the flame, and the blue calcium uie is often seen widened 

 when the II and K lines of calcium are not seen at all In consequence of the 

 numerous chan<.es in spectra at difierent temperatures, Mr. Lockyer has advanced 

 theTd^ that ?he molecules of elementary matter are continually being more and 

 morrbroken up as their temperature is increased, and has put forward the hypothesis 

 Sat the chemTcal elements with which we are acquainted are not simple bodies, but 

 Ire themselv™ compounds of some other more simple substances. This theory is 

 founded on Mr. Lock^-er's comparisons of spectra -%the maps of Ingstro^ 

 Youn- and others, in which there are coincidences of many of the short li^^e^ ot the 

 spec 1-a Jf different substances. These short lines are termed basic hnes,_ since hey 

 appeal to be common to two or more substances. They appear at the highest tem- 

 pKres when the longest hues of those substances and those which are considered 

 the te,st of their presence are entirely absent. . . , . , 



Mr Lockyer draws a distinction between weak lines, which are basic, ...which 

 woidd pennanently exist at a higher temperature in a more elementary stage, and 

 Te we™ short lines which wonld be more strongly present at a lower tempe- 

 ?itu-e ill a 1 e complex stage of the molecules. Thus m lithium, the i^d 1 ne 

 is a low temperature liSe, and the yellow is feeble ; at a ^jf - temperatu^^^^^^^^^^ 

 line is weak; the yellow comes out more strongly, and the blue line appears, at a 

 SerSSi-atuie still, the red line disappears, and the yellow dies away; whilst 

 SS^terpSui of the sun the violet lithium line is the only one which comes 

 out stiSX These effects are studied by first producing the spectrum of the 



:ltancet' the Bunsen flames, and observing the .^^^-f^ ^^JJ ^f^ J^^"^^^^^ 



„ .,,„-,.i. fiivniirrli tlip flaiiip • tlius. lu maorncsium a wide triplet oi set oi luiets 

 passing a spark tniougn tne name , mus, m i" o i i. /vi v2 „r,A l^•^^ of thp same 



lines f 5909-8 b^ andb-) is changed into a narrow triplet (V, b , and b ) ot tne same 

 charactr We have here what some observers regard as a recurrence of the same 



Skver holds that the different laye/s of the solar atmosphere may be regardec a^s 

 a serLs of furnaces, in the hottest of which, A, we have the most elementary 

 Irms of ma ei Sable of existing only in its uncombined state ; at a higher and 

 cook teT B thK m of matter may form a compound body and may no longer 

 exit in If ee^state at the lower temperature ; as the cooler and cooler levels CD 

 and E are reached, the substances'^ become more and more complex and form 

 different combTnatSns, and their spectra become altered at every stage. Since the 

 succeSveTYersarenotatrest,but in a state of disturbance, we may, get them 

 somewhat mixed, and the lines at the cooler levels D and E may be associated with 

 SriresoT he' hotter levels; these would be basic or ^o^^^^f «* ^^j, ^^ ^^^ 

 spectra of two different compounds which exist at the cooler levels D and E. We 

 SI eJen^et lines which are not present in the hottest furnace A coming into 

 Stence^?he ines of compounds in B or 0, and then extending among the mes 

 SoSrinVto more complex compounds which can only exist at a lower temperature 

 ltn° hey niTgl" be p^resent I coincident weak lines in the ^P-t-ot several 

 Compound bodies. Thus Mr. Lockyer regards the calcium hues H ^nd K of the 

 sSsSectrum as evidence of different molecular groupings of more elementary 

 bodL In the electric arc with a weak current the single hne 4226 of calcium, 

 which is easlv reversed, is much thicker than the two lines II and Iv ; but the three 

 Snes are equally hick 4ith a stronger current, and are all reversed. With a spark 

 ftom a arge S and using a condenser the line 4226 disappears, and H and K are 



