Dec. 23, 1881.] 



KNOWLEDGE 



151 



Ofb^NCE 



FLAiNLfWoRDED -EXACTOTESCRIBED 



LOXDOX: FRIDAY, DECEMBER 23, 1881. 



CONTENTS. 



Ih- So-callcl Elements. Bv Prof, 

 r. A. Yt. 



131 



\Vlml is a Grape ? By Grant All. 

 siudies of Volcanic Action. Part II. 



Bv G. F. Rodwell 151 



Our rnbiddon Gnesls, Part II. 



Ht Dr. Andrew Wilson. F.R.S.E. l"^ 

 - 'ids, Liquid'', and Gaseg. Part T. 



rvW. Mnttieu Williams l.iC 



■ li^olution of Man _ l.'w 



I .Met's Seismometers {Tlliutmle't) 15" 

 T\w Faure Accumnlator. By W. 



I.vnd 13S 



Tlr.' Wyandotte Indians. By Miss 



.U W. Buckland loS 



TAOB 



Xote on the Splieoridal Slate. By 

 W. F. Barrett, F.li.S.E 159 



An Instrument for Ih-awint: the 

 Conic Sections. By Thomas Thorp 

 {lUiutratei) 160 



CORBESPOKDENCB , 160 



The Northern Skies in Decembor— 



(niuslraled) 161 



~ ■ 167 



RcpUea to Queries 



Xotes on Art and Science i... 169 



Our Mathematical Column 169 



Our Chess Column 170 



Our Whist Column 171 



Answers to Corrospondejits 172 



THE SO-CALLED ELEMENTS. 



Bv Prof, G. A. Young. 



Y own spectroscopic oliservations bearing on the 

 elcraeutary constitution of matter have not been very 

 mimerous or important, as compared with those of several 

 other investigatoi-s ; at the same time, they liave liad a 

 place in the discussion of the subject, and I cheerfully 

 comply with the Editor's rciquest to present a short account 

 of them to the readers of Knowledge. 



The idea that our so-called chemical elements are not 

 i-.all}' and absolutely elementaiy, but arc built up either of 

 atoms of some one universal matter, or of some few sub- 

 stances of a higher order of simplicity than themselves, is 

 neither improbable nor new. As Mr. Ijockyer points out, 

 I?umas in 1836, and Brodic in 1867, not to mention 

 others, have expressed themsehes very clearly in this 

 direction. 



But the present special Uitcrest in the question is due 

 mainly to Mr. Lockyer himself, who, in 1873, by an able 

 ind extensive induction from all available spectroscopic 



-i-rvationR, put the theory in a new light, and brought 



,'cther in its favour a great Ijody of evidence, to which 

 he has since then been continually adding. 



In tlic discussion of the subject, certain observations of 

 ray o^Ti have been made use of along with those of others, 

 always, I believe, with due and courteous acknowledgment, 

 and in a manner to which I have nothing to object, unless, 

 perhaps, that too much weight ha.s sometimes been gi\en 

 to certain numbers which I had assigned only provisionally, 

 and on a very short experience. 



These observations of mine were for the most part made 

 iii.July and August, 1872, during a six weeks' stay at 

 Sherman, the summit station of the Fnion Pacific Iltul- 

 road, at an elevation of 8,300 feet above the sea Isvel. 

 Tlie party was connected with the organisation of the 

 Coast Survey, and the results of the observations are 

 published in the " Coast Survey Report " for 1872. 

 _ The work consisted mainly in a record of the bright 

 Imes observed in the spectrum of the solar prominences, 

 and of the lines which were widened or otherwise modified 

 in the spectra of sun-spots. The resulting catalogues con- 



tained respectively 273 and 1-56 lines, but only that portion 

 of the .spot spectrum lying between B and b was included. 

 In these catalogues an attempt was made to assign numbers 

 indicating the relative intensity of the lines and their pro- 

 portional frequency of appearance ; but, as has been inti- 

 mated, the number of observations was far too small to 

 make such figures very reliable. 



Several remarkable facts were, however, very clearly 

 bi-ought out. In the first place, those lines of the spectrum 

 of any particular substance (iron, for instance) which 

 are most conspicuously and frequently seen, as brir/hl 

 Imes in the spectrum of the chromosphere and prominences, 

 are not generally the same as those most notably widened 

 in the sun-spot spectrum ; nor are either of these the ones 

 wliich are most prominent in the ordinary solar spectrum 

 or in the spectrum of the electric arc. Each of these four 

 spectra selects and emphasizes, so to speak, its own set of 

 so-called iron lines, and the same is true of other sub- 

 stances. Second : Where distortions and disturbances of 

 the lines indicated rapid motions of the solar gases alonn- 

 tlie line of sight (approaching to, or receding from, the 

 observer), it frequently happened that neighbouring lines, 

 due to different substances, were differently affected ; 

 certain lines of iron might be violently displaced, while 

 the adjacent lines of calcium were not disturbed at all. In 

 some instances, also, it was noticed that lines, given in 

 the maps as belonging to the same element, and actually in 

 the same field of view, behaved very differently ; one of 

 them indicating violent motion of the gas concerned, the 

 other showing nothing of the kind. But this observation 

 seemed at the time so strange, and the instances were so 

 few, that, in view of the possibility of mistake, it was not 

 thought best to publish it without further confirmation, 

 which it has since abundantly received. 



In the third place, and most important, it was found 

 that a very surprising proportion of the lines con.spicuous 

 in the spectra of the chromosphere and sun-spots are lines 

 which on the maps are given as common to tlie spectra of 

 two or more elements — lines which Mr. Lockyer has since 

 designated by the term "basic." Nearly one-third of all 

 the lines in the two catalogues belong to this class, while in 

 the ordinary solar spectrum, they amount to less than four 

 per cent, of the whole. 



At that time the prevailing opinion, I think, was that 

 these common lines were due to impurities in the .specimens 

 of the metals employed in mapping the spectra. If, for 

 instance, in making a map of the spectrum of iron, the bits 

 of iron used for the electric spai'k to play between (techni- 

 cally the " electrodes '') were slightly contaminated with 

 calcium, then we should expect the more important lines of 

 calcium to appear faintly in the map of the iron spectrum ; 

 and the difficulty of getting perfectly, or even " spectro- 

 scopically," pure iron is enormous. 



It was evidently possible also that these coincidences 

 might be such only in appearance— cases of accidental 

 juxtaposition too close for resolution by the existing 

 spectroscopes. 



But the instances were so numerous, tliat, on the whole, 

 it looked more prolialile that the true explanation of such 

 coincidences would be found in something common to the 

 elements concerned; and at the time, I thought a siuiiiarity 

 of molecular structure! more likely than a community of 

 substances. It appeared very possible that metals so closely 

 alike in many of their properties as iron and chromium, 

 for instance, might owe this general resemblance to such a 

 likeness of their molecular arcliitecture as would also carry 

 with it certain identical vibration-periods. If atoms are 

 vortex-rings in a continuous medium, the likeness might Vjo 

 in the size and configuration of the rings ; if, on the other 



