October 24, 1901J 



NATURE 



629 



have been atoms very nearly at rest ' or quite at rest ; 

 more densely distributed in some places than in others, of 

 infinitely small average density through the whole of 

 infinite space. In regions where the density was then 

 greater than in neighbouring regions, the density would 

 become greater still ; in places of less density, the density 

 will become less : and large regions will quickly become 

 void or nearly void of atoms. These large void regions 

 would extend so as to completely surround regions of 

 greater density. In some part or parts of each cluster of 

 atoms thus isolated, condensation would go on by motions 

 in all directions not generally convergent to points, and 

 with no perceptible mutual influence between the atoms 

 until the density becomes something like lo"'' of our 

 ordinary atmospheric density, when mutual influence by 

 collisions would begin to become practically effective. 

 Each collision would give rise to a train of waves in ether. 

 These waves would carry away energy, spreading it out 

 through the void ether of infinite space. The loss of 

 energy thus taken away from the atoms would reduce 

 large condensing clusters to the condition of gas in equi- 

 librium - under the influence of its own gravity only, or 

 rotating like our sun or moving at moderate speeds as in 

 spiral nebulas, &c. Gravitational condensation would at 

 first produce rise of temperature, followed later by cooling 

 and ultimately freezing, giving solid bodies, collisions 

 between which will produce meteoric stones such as we 

 see them. We cannot regard as probable that these 

 lumps of broken-looking solid matter (something like the 

 broken stones used on our macadamised roads) are prim- 

 itive forms in which matter was created. Hence we are 

 forced in this twentieth century to views regarding the 

 atomic origin of all things closely resembling those pre- 

 sented by Democritus, Epicurus, and their majestic 

 Roman poetic expositor, Lucretius. 



THE CHEMISTRY OF THE CYGNIAN STARS 



AND BASIC ROCKS. 

 T HA\'E recently received from Prof. Suess the follow- 

 ■•■ ing important letter, with a request that it should 

 be sent for publication in Nature. It is obvious that 

 Prof. Suess' striking generalisation will lead to many 

 interesting inquiries ; for my part I shall lose no time 

 in making the additional researches for which he asks. 

 The results on the chromosphere spectrum obtained 

 during the eclipse of 1S98 now being published by the 

 Royal Society will also, I think, throw some light on 

 the question. Norji.^n Lockver. 



Vienna, October 7. 



■My Dear Sir. — In reading your highly suggestive 

 and instructive book on Inorganic Evolution and your 

 last papers in Proc. Roy. Soc. and Nature, I was 

 struck by what you say on the spectrum of a Cygni, 

 and beg leave to submit a question. 



I believe I am in accord with the best masters of 

 geology of our time in regarding our earth as a ball of 

 NiFe, surrounded by a silicious slag. This slag has 

 parted (or has differentiated) into two zones, one richer 

 in SiAl and felspathic minerals (trachyte, granite, &c.), 

 and the other richer in SiMg (peridotite and serpentine, 

 olivine-rock, Iherzolite, dunite, kyschtymite, &c.), and 

 both extremes are united by a host of intermediate rocks. 

 The SiAl group is lighter, exterior and partly used up in 

 forming sediments, the SiMg group is lower or interior 

 and related by the universal occurrence of traces of Ni, 

 and by other features, to NiFe. They are, in fact, the 

 acid and the basic group of Bunsen, Durocher and all 

 their followers. 



^ "On Mechanical Antecedents of Motion, Heat and Light"; Brit. 

 Assoc. Rep., part ii. 1854 ; Edin. New Phil. Joui\ i. 1855 ; Comptes reiidiis, 

 .vl. 1855 ; Kelvin's " Collected Math, and Phys Papers," vol. ii. art. l.^ix. 



- Homer Lane, .American Journal of Science, 1870, p. 57 ; Sir W. 

 Thomson, Phil. .^!ag., March 1887, p. 287. 



Now Prof. Vogt, of Christiania, has in a series of 

 remarkable papers shown that the metallic ores acconi- 

 panying the SiAl rocks are different from the ores of the 

 SiMg group, and that certain metals are very charac- 

 teristic for each group, and more especially for certain 

 ultra-basic rocks. The same author has, in Zcitschrift 

 fiir prakt. Geolog., 1898, p. 326, given a list of the 

 elements characteristic for the ores of each group. This 

 list does not regard the rocks themselves, but only the 

 ores, and I have thought to give a better approach to the 

 sum of occurring elements by introducing a few trifling 

 changes, viz. in noting Si, Al and also Cu on both sides, 

 and adding C, according to South African experience, to 

 the basic list. 



(i) Acid rocks : Si, K, Li, Be, Al, W, U, Ce, V, Cu, Sn, 

 Zr, Th, B, F. 



(2) Basic rocks : Si, Ca, .\\, Ba, Sr, Mg, Fe, Mn, Ni, Co; 

 Cr, Cu, V, Ti, Ft and allied metals, C, P, S, CI. 



(The most characteristic members are italicised.) 



The ultra-basic rocks have their exact equivalent in 

 the meteoric stone of Chassigny. 



The great number of analyses of meteor irons by 

 Cohen recently published by the Berlin Academy gives 

 Fe, Ni, Co, Cr, Cu, C, P, S, CI, joined in single cases by 

 SiOj, MgO and CaO. You know that Davison also 

 found Pt and probably Ir. It is Vogt's list of metals 

 from basic rocks. 



Angstrom's older list of Fraunhofer lines gives H and 

 Na, and besides these Ca, Ba, Mg, Fe, Mn, Ni, Co, Cr 

 and Ti. 



These are all elements accompanying the basic rocks, 

 and although your own later observations show the 

 existence of lines of some of the characteristic metals of 

 the acid series, such as K, Li, U and Ce, I must con- 

 clude that Vogt's basic list is more distinctly repre- 

 sented in the sun's absorbing layer. 



Now your comparison of the sun's chromosphere 

 with the enhanced lines of a Cygni gives a quite similar 

 list for a Cygni : Mg, Ca, Fe, Ti, Mn, Ni, Cr, V, Cu, Sr, 

 and the impression is that the metallic vapours, which 

 accompanied the intrusion of ultra-basic rocks and 

 sometimes, as in Norway, gathered as ores at the cir- 

 cumference of these ultra-basic intrusive rocks, must 

 have been of a remarkable likeness to those of the sun's 

 chromosphere and of a Cygni. 



One might even be induced to go a step further. 

 Among the ores of the ultra-basic rocks Vogt distinguishes 

 two varieties, which he calls the oxydic, characterised by 

 the prevalent occurrence of Ti, and the sulfidic ores 

 (nickeliferous pyrrhotite). And I find that you remark 

 7 Cygni to be distinguished by the prevalence of Ti, 

 a Cygni by Fe, Cr and Ni, and h Orionis by Si and Mg 

 — corresponding to these two varieties of ores and to 

 the intrusive rock SiMg {ji Orionis being the country 

 rock of the ores). But I am very far from proposing any 

 conclusions whatever in this imperfect state of know- 

 ledge, and only venture to point out the interest which 

 is attached to the examination of a number of elements 

 named by you on p. 58, " Inorganic Evolution" — which 

 have only been investigated by you with lower dispersion, 

 and which embrace several typical representatives of the 

 acid series — and to the special research of metals like 

 W, U, Ce, Sn and others of this series. 



The question, which I take leave to submit, is : Have 

 we indeed to suppose that metallic vapours answering to 

 metals from acid rocks are less represented in the sun 

 and stars than those from basic rocks, or is it some 

 secondary cause, the nature of their spectrum or other, 

 which gives the present seeming prevalence to the metals 

 from basic rocks ? 



I beg you, dear sir, to accept the expression of my 

 highest esteem. 



Yours most respectfully, 



Edw. Suess. 



NO. 1669, VOL. 64] 



