July 8, 1880] 



NA TURE 



'Z^ 



that M. Soret has priority in the discovery of these new earths, 

 since the absorption-spectrum of holinium coincides exactly with 

 that of the earth X., and thulium is probably the same as the un- 

 named earth of Soret or the phillipium of Delafontaine. The 

 existence of these three earths in the yttria group is also acknow- 

 ledged to some extent by Marignac ; ^ it may therefore be con- 

 sidered tolerably certain that these new earths are really contained 

 in this group, ^^'hatever names they may ultimately receive. In 

 connection with this we must not omit to mention the investi- 

 gations of M. Lecocq de Boisbaudran. He has confirmed- the 

 results of MM. Soret and Cleve concerning the three earths 

 mentioned above, and even thinks that he has obtained sufficient 

 evidence of a fourth, named samarium. He admits, however, 

 that its separation is too tedious to allow of its extraction in a 

 state approaching purity. 



In operating upon the mixture of "earths formerly known as 

 erbia Nilson was able to separate, besides the earths giving 

 absoiption-spectra and besides ytterbia, another white earth, 

 whose salts gave no spectrum and which possessed a low atomic 

 weight (about 45). The new metal contained in this earth he 

 named scandium,'^ and he states that it is distinguished by a 

 special spark-spectrum. These results have been confirmed by 

 Cleve,'' and he has remarked that many of the properties of this 

 scandium agi'ee clo?e!y with tlie metal whose existence was 

 predicted by Mendelef under the name of ekabor. 



The following are a few of the more striking of these 

 resemblances : — 



Ekabor Scandium 



At. wt. = 44 At. wt. = 45 



Only oxide = Eb^Oj Only oxide = ScjOj 



The oxide is white, infusible, and nearly allied to yttria. 

 tS.G. of oxide = 3-5 S.G. of oxide = 3-8 



On the other hand it is difficult to understand how a metal 

 with such a low atomic weight could remain associated with 

 others possessing atomic weights three or four times as great 

 throughout the long process of fractional separation. According 

 to all analogy with yttrium, terbium, and erbium, it ought to 

 remain with the first of these. The folbning table of the 

 metals of the yttrium group will illustrate the present state of 

 our knowledge with regard to them. The atomic weights are 

 calculated on the supposition that their oxides are of the general 

 formula MoO^ — those in italics give distinct absorption-spectra : — ^ 



It must also be remarked that Delafontaine has suspected that 

 the didymia obtained from cerite differs from that from 

 samarskite, although Lecoq de BDisbaudran and L. Smith have 

 since shown that the absorption-spectrum of the didymia salts 

 may be considerably altered by making the solutions strongly 

 acid, &c. And it is of course open to question whether some of 

 the spectroscopic diflerence^ ascribed to different metals may not 

 be due to differences in the concentration, acidity, &c., of the 

 solutions employed. 



It only remains to mention the newly-discovered metals — 

 iwrwegium and vesbitim. The former was announced to English 

 chemists some twelve months ago by Dr. T. Dahll.^ It is, he 

 says, a white metal, allied to copper in many of its properties, 

 but with a melting-point of about 350° C, and a specific gravity 

 of about 9'4. Its atomic weight would lie between 141 "5 and 

 lSO"6. The latter metal (vesbium) hns been stated by M. A, 

 Scacchi to be present in a green incrustation found on Vesuvius 

 in the fissures of the eruption of 1631. It is, he says, present in 



* A rch. des Sci„ j^hys. et nat. {loc. ctt.). 



'-' Comp. Rend.t vol. bcxxlx, pp. 212 and 516. 



3 Ibid., voL Ixxxviii. p. 645. * Comp. Rend., vol. Ixxxix. p. 419. 



5 Lawrence Smith's Mosandrum is not included in this list, since its 

 existence is so improbable. 



6 Chem. Nevjs, vol. xl. p. 25. 



the shape of a red metallic acid, giving colourless salts with the 

 alkalies. Many of its properties agree with those of molybdenum 

 or vanadium, particularly the latter, though M. Scacchi believes 

 that both these metals are absent. Of numerical data only the 

 proportion of silver in the silver salt is given. Thi., is stated to 

 be 4S'8, while for the corresponding vanadium salt it would be 

 52'i, a coincidence too close to be disregarded. ^ Up to the 

 present we are without any confirmation of the existence of these 

 two metals, and we cannot do otherwise than suspend judgment 

 on them for a time. 



Indeed the scepticism which the chemist, in common with 

 other scientific men, ought to practise cannot be too strongly 

 insisted upon. No discovery of such importance as that of a 

 new element should be generally accepted until it has been sub- 

 mitted to a series of rigorous confirmatoi'y tests. It is obviously 

 so much better to defer definite judgment until sufficient facts have 

 been collected than to accept a hasty conclusion, probably based 

 only upon one or two anomalous reactions. How often it happens 

 that the chemist describes a reaction not as he saw it, but as he 

 tliought he saw it, or as he hoped to see it ! Even in cases where 

 the reaction possesses some peculiarity too little attention is often 

 paid to the effects which] even traces of other substances may 

 produce, or to any extraordinary conditions under which the 

 experiment may be made, and the chemist at once imagines that 

 he has discovered a "new element." Time ialone will prove 

 how many of the fourteen substances enumerated above w'ill 

 pass the ordeal of further and perhaps more rigorous investigations. 



T. S. HUMPIDGE 



Since writing the above M. Marignac has published an account 

 of some investigations on the earths contained in samarskite.^ 

 He divides these earths into four groups, according to their 

 solubility, in a saturated solution of potassium sulphate : — 



(i.) Those earths soluble in less than 100 parts of the solution, 

 (ii.) Those soluble in 100 to 200 parts. 



(iii.) Those only slightly soluble. 



(iv.) Those insoluble. 



Group (i.) contains only well-known earths, and particularly 

 yttria and terbia. Their equivalent was always below 119 

 (oxide = MO). Group (ii.) consists of earths with an equivalent 

 between 119 and 115. It contains traces of the preceding and 

 following groups, but principally consists of a pale yellow earth 

 with an equivalent of about I20'5, and without any absorption 

 spectrum. This earth he provisionally calls Yo ; its properties 

 do not agree with those of any of the others of this group men- 

 tioned above. Group (iii.) contain? a considerable quantity of 

 terbia and didymia, together with a colourless earth yielding an 

 absorption-spectrum agreeing with that of Delafontaine's decipia, 

 or better with that of Lecoq de Boisbaudran's samaria. This 

 earth he calls Yj8, and he is of opinion that decipia, samaria, 

 and Y^ are practically one and the same earth. The equivalent 

 he makes li5'6 (oxide — MO), which would give an atomic 

 weight of I49'4 (oxide MoOj). Group (iv.) consists principally 

 of didymia, together with considerable portions of the other 

 earths, which it is almost impossible to completely separate. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE 



Oxford.— At Trinity College one Millard Scholarship, 

 tenable for four years during residence, and of the annual value 

 of 80/. without limit of age, will be awarded in October next 

 for proficiency in natural science if any candidate of sufficient 

 merit offers himself. The subjects of examination will be che- 

 mistry and physics. Candidates may also offer mathematics, if 

 they wish to do so, and give notice a week before the examina- 

 tion. Special weight will be attached to excellence in one or 

 two subjects, rather than to a less thorough knowledge of all. 

 The scholar elected will not necessarily be required to commence 

 residence immediately. The same papers will be set in chemistry 

 and physics as in the examination for the Natural Science 

 Scholarship at Exeter College. Every candidate will be con- 

 sidered as standing at both colleges, unless he makes a statement 

 to the contrary on entering his name. Candidates are requested 

 to state which college they would prefer in the event of their 

 being elected at both colleges. The president will receive the 

 names of candidates, and their testimonials of character, on 

 Wednesday, October 13, between 8 and 9 p.m. 



' Ber. d. deut. ch. Cisell., vol. xiii. p. 2S0; Nature, vol. xxi. p. 420. 

 = Compt. Rend. vol. xc, p. 899. 



