358 



NA TURE 



{^August II, 1887 



Similar observations were made upon the nitrates derived from 

 the other minerals above mentioned, the actual wave-lengths 

 being in every case determined so that the position of the lines 

 can be open to no doubt whatever. 



Ten years ago the erbium earths were considered as the oxide 

 of a single element, but we now know that they consist of the 

 oxides of scandium, ytterbium, thulium, erbium, terbium, hol- 

 mium, and yttrium. Out of the rich data furnished by the 

 present observations the observers believe they can prove that all 

 those erbium*earths whose nitrates give absorption-spectra are 

 not oxides of simple bodies, but mixtures of the oxides of various 

 new elements. Yttrium, as before mentioned, has already been 

 shown by Mr. Crookes to consist of five constituents, and it will 

 be interesting to see what light the workers in Stockholm have 

 thrown upon the nature of some of the remainder. 



M. Lecoq de Boisbaudran showed that by fractionation of 

 Soret's X two new substances were arrived at, which he named 

 holmium and dysprosium, but these are now shown to be them- 

 selves compound, for one part of dysprosium is not present in 

 thorite of Brevig or cerite of Bastnas, although present in the 

 mixture called X ; in fact, de Boisbaudran's dysprosium lines 

 Dya, Dy3, and Dy5 are found to belong to three different 

 elements; and the other constituent, the holmium of de Bois- 

 baudran, is probably made up of no less than four distinct 

 components. 



As the introduction of fresh names is rapidly increasing the 

 difficulty of work in this direction, Krliss and Nilson prefer to 

 simply label the components by affixing the letters of the Greek 

 alphabet to already accepted symbols. The metal called by 

 Soret X is therefore constituted as follows : — 



Fergusonite of Ytterby and 

 euxenite of Arendal and 

 Hittero contain 



Xa, X)8, X7, X5, Xe, XC, Xrj. 



We are now accustomed to distinguish as erbium that body 

 whose nitrate solution exhibits, in addition to a large number of 

 lines in the violet and ultra-violet, two principal lines of wave- 

 lengths 523 'I and 654-7 respectively, of which the former is the 

 most intense. But in euxenite of Hittero much greater differ- 

 ence is shown, one being extremely strong, while the other is 

 barely visible ; therefore here again the observers consider 

 themselves in face of at least two elements, Era and Er^, one 

 giving 523-1 and the other 654-7. Moreover, they have succeeded 

 in separating the two almost completely by a method of fraction- 

 ation similar to that employed by Mr. Crookes. 



Cleve, in 1879, gave the name of thulium to the metal whose 

 oxide formed the strongest base present in the mixture of erbium 

 earths ; and its salts, according to Thalen, exhibit two absorption- 

 bands, 684-0 and 465-0, of which the former is the most intense. 

 Again, the variations are found to be too great for the supposi- 

 tion of a single earth to be tenable, one line being entirely 

 absent in fergusonite and thorite of Arendal, while the other is 

 strong ; hence thulium must also consist of two ingredients, Tma 

 and Tmj8. 



The observations with regard to didymium are all the more 

 interesting as entirely confirming Mr. Crookes's statements, and 

 Drs. Kriiss and Nilson even go further in proving either that 

 our interpretations of the indications of spectrum analysis are 

 grossly wrong or that didymium is composed of not less than 

 nine distinct elements. Dr. Auer von Welsbach's symbols for 

 praseodymium and neodymium, the two constituents of didymium 

 which he actually separated, are discarded, and the same nomen- 

 clature adopted as in the case of holmium. 



Dia 

 Dij8 

 Di7 

 Di8 

 Dif 

 DiC 

 Dir; 

 Die 

 Di. 



Characteristic line in 

 absorption-spectrum of 

 nitrate solution. 

 728-3 

 679-6 



579-2 and 575-4 



521-5 

 512-2 

 4820 

 4690 

 445"i 

 444*7 



The name samarium was given by M. de Boisbaudran to an 

 element identical with Marignac's Y/3, an ingredient of the old 

 didymium. The nitrate of this metal gives seven absorption- 

 bands according to Thalen, but it is surprising that in thorite 

 and euxenite of Arendal the line 416-7 is tolerably strong, and 

 even very strong, without another samarium line to be seen in the 

 spectrum ; the conclusion is inevitable that there must be in this 

 substance a constituent whose nitrate gives the line 4167, and 

 to this the name Sma is given, all other samarium lines being 

 provisionally supposed to belong to Sm/3. 



The main result of this splendid work, therefore, appears to 

 be that, instead of holmium, erbium, thulium, didymium, and 

 samarium, we must, if we follow Kriiss and Nilson, recognize 

 the existence of at least twenty- two new elements, the fate of 

 some of which may be, in the near future, to be subjected to still 

 further subdivision. If we add to these the results previously 

 obtained by Mr. Crookes with respect to yttrium, the astound- 

 ing revelation is presented to us that instead of six we find 

 ourselves in face of twenty-seven, or a clear gain of at least 

 twenty-one new elements. 



But now comes the vital question — Are these really new 

 elements, or are they only different molecular aggregations of the 

 atoms of a few, as suggested by Mr. Crookes? It certainly 

 seems very remarkable that so large a number of elements should 

 be crowded together about the central series of the periodic 

 system, and we appear to have a repetition of the same pheno- 

 menon, in a much intensified degree, as obtains in the cases of 

 nickel and cobalt, rhodium, ruthenium, and palladium, and of 

 iridium, osmium, and platinum. It may, however, be interest- 

 in this connexion to remember that this precise state of things 

 was predicted by Mendelejeff himself {Ann. Chem. Pharm. 

 Suppl. 8, p. 158), and in no way militates against the new 

 element theor}'. Kriiss and Nilson, rather than be obliged to 

 have recourse to the introduction of new or auxiliary theories [of 

 spectrum analysis, prefer to rest upon the simpler and apparently 

 more straightforward assumption that these substances, whose 

 nitrate spectra show such marked differences, are indeed bond fide 

 new elements. The accuracy of this view will doubtless be 

 strongly contested, but in any case the result appears likely to 

 be equally striking ; for, if future work shows its want of accord- 

 ance with facts, then an entirely new field of research has been 

 opened, and the generally accepted ideas of the structure of 

 matter must of necessity undergo a complete metamorphosis. 



A. E. TUTTON. 



SOCIETIES AND ACADEMIES. 

 London. 



Royal Society, June 16. — "On Figures of Equilibrium of 

 Rotating Masses of Fluid." By G. H. Darwin, M.A., LL.D., 

 F.R. S., Fellow of Trinity College and Plumian Professor in 

 the University of Cambridge. 



The intention of this paper is, first, to investigate the forms 

 which two masses of fluid assume when they revolve in close 

 proximity about one another, without relative motion of their 

 parts ; and secondly, to obtain a representation of the single 

 form of equilibrium which must exist when the two masses ap 

 proach so near to one another as just to coalesce into a singl 

 mass. 



When the two masses are far apart the solution of the prob- 

 lem is simply that of the equilibrium theory of the tides. Each 

 mass may, as far as the action on the other is concerned, be 

 treated as spherical. When they are brought nearer to one 

 another this approximation ceases to be sufficient, and 

 departure from sphericity of each mass begins to exercise 

 sensible deforming influence on the other. 



The actual figure assumed by either mass may be regarded j 



