June, 1903.1 



KNOWLEDGE. 



129 



Physics Observatory, South Keusingtuu, gives an excellent 

 idea of the Arcturian type of sjiectnun, with its numerous 

 well-defined lines. When the spectrum is carefully dis- 

 sected, it is found that it reproduces the solar spectrum 

 with remarkable fidelity, even in minute details. In 

 rig. 8 we have an enlarged view of the small part of the 

 spectrum between wave-length 4000 and wave-length 

 4150, which should be compared with the photographs of 

 the same part of the solar spectrum previously given in 

 Figs. 2 and 3 (KNOWLEDflE, February, 1903). Comparing 

 it with the solar spectrum photographed with moderate 

 dispersion (Fig. 2), the similarity is especially manifest, 

 and the stellar spectrum only dift'ers from the more perfect 

 photograph of the solar spectrum shown in Fig. 3 in 

 compounding groups of close lines into single lines. There 

 is no reason to supj)ose, however, that such groups would 

 fail to show the separate components if instruments of 

 sufficient dispersion could be employed on the stars. 



It is evident then that the sun is tlie best example of 

 the Arcturian group of stars, and that an interpretation of 

 its spectrum will also hold good for the other stars of the 

 same type. As already exp)lained, this interpretation is 

 based on matching the solar lines with lines of terrestrial 

 substances, preferably by photographic methods, and it is 

 then found that most of the lines correspond in wave-length 

 and relative intensity with those which appear in the arc 

 spectra of various metals. The results of such comparisons 

 are best exhibited by tabulation of wave-lengths, and the 

 accompanying brief extract from Rowland's great table of 

 something over twenty thousand solar lines will indicate 

 the nature of the results obtained by the use of instru- 

 ments of very great dispersion. 



(The table should be compared with Figures 2, 3, and 8.) 



Wave-lengtli. j Substance. Intensity. Ij Wave-length. Substance, Intensity 



It will be observed that many of the lines, chiefly of 

 small intensity (as indicated by the intensity numbers, 

 0, 0<>, 000), have not yet been identified, but, as the 

 possible comparisons are by no means exhausted, it is uot 

 vet to be concluded that such lines have no terrestrial 

 e<iuivalents. The elements so far recognized as contri- 

 buting to the solar spectrum in its entirety are as 

 follow : — 



In addition there is evidence which suggests the possible 

 presence of the following elements : — 



Beryllium. Lead. Ruthenium, 



Didymium. Lithium, Tantalum, 



Erbium. Mercury. Thallium. 



Germanium. Niobium. Thorium. 



Glucinum. Osmium. Tin. 



Indium. Platinum. Tungsten. 



Iridium. Rhodium. Uranium. 



Besides these, though not revealing their presence in the 

 Fraunhofer spectrum of dark lines, helium and the hypo- 

 thetical " corouium " are important constituents of the 

 solar (•lir()uiosj)lR're ;nid curona respectively, and there can 

 l:)e little (Iciubt tjiat iudications of these gases would also bo 

 found in the Arcturian stars if we were near enough to 

 disentangle the spectra of the separate parts of the stars, 

 as we are able to do in the case of the sun. This example 

 forcibly illustrates the important truth that the investi ■ 

 gation of such a spectrum as that of an Arcturian star can 

 only directly reveal the chemical composition of the par- 

 ticular part of the star which by its absorption produces 

 the dark lines. The spectrum indicates what is certainly 

 present, but gives no evidence as to what is really absent. 



The general result of this analysis of the solar spectrum 

 then is to show that the elements entering into the 

 composition of stars of the second t3^pe are not essentially 

 different from those composing the earth, and this 

 ol)viously is a conclusion of the first importance in con- 

 siderations relating to the plan of the universe. Its 

 significance can be partially realised when the number and 

 distribution of the stars are investigated. There is already 

 evidence that some thousands of stars have spectra of the 

 second type, but dealing only with those which have been 

 studied in detail, Lockyer finds that out of 470 stars* 

 (including all the stars down to magnitude 3o in both 

 hemispheres) there are 78 of the Ai'cturian and 56 of 

 the Aldebarian group ; that is, in more than a quarter of 

 the stars investigated there is direct evidence of the 

 predominance of solar and terrestrial matter. It is indeed 

 not too much to say that not only are the same substances 

 present in all these stars, but that they exist in the same 

 proportions throughout. 



A study of the (listribution of stars of this type is not 

 less suggestive. Some of the brighter members, according 

 to Lockyer, are as follow : — 



Aldebakian Stabs, .\BCTrRiAX Stabs. 



Y Andromedie. » Aquarii. 



Y Aquilae. « Arietia. 

 ^ Capricorni. a Aurig.'c. 

 » Hydnc. « Bootis. 



t Pegasi. ' "', «■- Capricorni. 



« Pho'nicls. " Cassiopei;v. 



n Serpentis. » Centauri. 



" Tauri. ^ Gcminorura. 



a Toucan i. y Lcouia, 



P UrsiB Minoris. y Poisei. 



« A'irginis. 3 Tirginis. 



These few exam])les will serve to show that the second 

 type stars surround us in every direction, and a fiirthei- 

 consideration of their distances, so far as they have been 

 ascertained, suffice at least to show the immensity of the 

 space through which terrestrial matter is dispersed. Re- 

 cently published determinations by Elkiu give the parallax 

 of a Tauri as '•10S>. of a Aurigw as 0"079. of /3 Geminorum 

 as 0''-05t), and of a Bootis as 0"02(i. corresponding re- 

 spectively to 30, 41, 59, and 125 hght-years. Taking 

 proper motion as an indication of stellar distance, the 



• "Catalogue of 470 of the Bright.er Stars, classified according to 

 their chemistry at the Solar Physics Observatory." (H.M. Stationerv 



Ollii-e. I'.'O-' i 



