;44 



NA TURE 



\_August 9, 1888 



absorbing molecules in the molecular groupings, which 

 other optical methods cannot indicate ; for the absorption 

 due to one molecule is independent of that of a neigh- 

 bouring molecule, while the phenomena of refraction only 

 show resultant effects. 



Further, as experiment shows that in most crystalline 

 substances the principal directions of absorption coincide 

 with the principal directions of optical elasticity, and as it 

 is probably right to assume that each molecule is subject 

 to the same laws as the whole of the crystal, there is no 

 reason to suppose that the directions of symmetry should 

 be different in the molecule and in the crystal, provided 

 the latter presents no optical anomaly. One can there- 

 fore assume that the principal directions of absorption in 

 the molecules themselves coincide with their axes 6f 

 optical elasticity, and that in mixed crystals the anomalous 

 directions of absorption indicate the direction of the optic 

 axes of the different absorbing substances. If this is 

 really the cause of the anomalies in the direction of certain 

 bands, each group of anomalous bands ought to belong to 

 different substances, of which the existence in the crystal 

 is thus revealed. 



To prove the truth of this beautiful theory, M. Becquerel 

 points out the significant fact that among the substances 

 which he finds to be characterized by anomalous bands, 

 several have been separated chemically into their 

 components. 



We have, therefore, in the observation of anomalous 

 directions of absorption a new method of spectral analysis, 

 a method of determining in a crystal the presence of iso- 

 morphous substances, optically dissimilar. We can even 

 go further still, and recognize the existence of different 

 substances, although they may not manifest anomalous 

 directions of absorption. For, suppose the same bands 

 are noticed to occur in the spectra of several crystals ; if 

 in one of these crystals two bands become maxima or 

 minima at the same time for the same direction of vibra- 

 tion, and if in another crystal one of them disappears for 

 the direction which renders the other a maximum, one 

 may conclude that the bands are due to two different 

 molecules. 



This new method of analysis appears to be specially 

 suitable for use in unravelling the mystery of the constitu- 

 tion of the rare earths. If, as seems now almost certain, 

 they consist of the oxides of a large number of element- 

 ary substances, the difficulty experienced in separating 

 them points to the fact that these constituent oxides must 

 resemble each other closely. It is therefore most prob- 

 able that their salts will be isomorphous, and the crystals 

 of these salts may consequently be expected to give 

 absorption spectra of great interest in the light of the 

 foregoing theory. M. Becquerel has therefore subjected 

 the crystalline salts of didymium to the test of experi. 

 ment, with the important result that several substances 

 have been detected which chemists have recently isolated 

 chemically ; and also new substances have been identified 

 as constituents, of which chemical methods have not as 

 yet revealed the presence. 



It will be remembered that Auer von Welsbach, by 

 fractional crystallization of the double nitrates of didy- 

 mium and ammonium, obtained two solutions — one 

 possessing a green colour, showing almost exclusively 

 the three bands X <= 482, 469, and 445, and which he 

 termed praseodymium ; the other a red solution, giving 

 the other bands of the didymium mixture except 

 X = 4755, which received the name neodymium. The 

 study of the absorption spectrum of crystals of sulphate 

 of didymium now shows that the two groups X = 483-6- 

 482*2 and X = 471*5-470, which have anomalous direc- 

 tions to a remarkable extent, are characteristic of 

 praseodymium, while most of the bands of neodymium 

 have directions quite different. Again, on examining 

 these same groups belonging to praseodymium in the 

 crystals of double nitrate of didymium and potassium, 



it is noticed that the bands which appear to have the 

 same principal directions in the sulphate have in the 

 double nitrate directions quite different, characterizing 

 two distinct substances. Later experiments by De- 

 marc_ay have indeed shown the possibility of chemically 

 isolating two constituents — one exhibiting the band X = 

 469, the other giving the bands of praseodymium. 



Hence the new method proves a most valuable test of 

 the accuracy of chemical work. In multiplying the ob- 

 servations,.]^!. Becquerel concludes that didymium is, as 

 expected, a mixture of a large number of substances, 

 chemically different ; among the identified constituents 

 are almost all that have been already chemically isolated, 

 and very probably others, notably one substance which 

 is characterized by the band X = 5717. 



A remarkable confirmation of this new law of crystal 

 absorption was obtained in the following way. When a 

 crystal of the sulphate or nitrate of didymium is dissolved 

 in water, the spectrum of absorption of the solution pre- 

 sents curious differences from that of the crystal. Certain 

 bands have remained permanent, but others are displaced, 

 and some have entirely disappeared. This is readily 

 explained if one admits that the crystal consisted of a 

 mixture of compounds unequally acted upon by water. 

 The most interesting fact, however, is that the bands 

 which manifest these variations are precisely those which 

 in the crystal present the anomalies. 



In conclusion, we see that by the employment of this 

 new method of analysis we are enabled, without destroy- 

 ing the crystal, as is necessary in chemical analysis, to 

 recognize the presence of different chemical molecules ; 

 and as we obtain three distinct spectra from the three 

 directions of optical elasticity, the method is one of 

 extreme sensibility. Every investigator likes to see his 

 work confirmed, and in this most difficult field of the 

 rare earths we cannot have too many confirmations. The 

 more points of the compass from which we approach it the 

 better, for we are sure then of surrounding and finally of 

 grasping the truth itself, in all its grand simplicity. 



A. E. TUTTON. 



THE NE W VEGE TA TION OF KRAKA TA O. 



'THE great volcanic eruption of Krakatab in August 

 -*■ 1883 will be fresh in most memories. It was at 

 one time reported that the island had totally disappeared, 

 but this was not so. Previous to the eruption, however, 

 it was covered with a luxuriant vegetation, no trace of 

 which existed after the event. 



Dr. M. Treub, the Director of the Botanic Garden at 

 Buitenzorg, Java, when at Kew last year informed the 

 writer that he had visited the island the previous year, 

 and intended publishing the results of his botanical 

 investigations. This he has now done, 1 and as the deri- 

 vation of insular floras is a subject of great interest to 

 many persons, the substance of Dr. Treub's communica- 

 tion deserves a place in. Nature. 



The existing portion of Krakatab Island is about three 

 miles across, and has the form of a mountain rising out 

 of the sea. On one side it is nearly perpendicular almost 

 to the summit of the peak, which has an altitude of about 

 2500 feet, and on the other it presents a steep slope. It 

 is situated ten miles distant from the Island of Sibesie, 

 the nearest point where there is terrestrial vegetation ; 

 twenty miles from Sumatra, and twenty-one miles from 

 Java. Verlaten and Lang Islands, lying much nearer 

 Krakatab, were utterly desolated and denuded of their 

 vegetation by the great catastrophe, and were still 

 absolutely bare in 1 886. 



With regard to the total destruction of vegetable life in 

 the island, Dr. Treub asserts that there can be no doubt : 



1 Annates du Jardin Botanijne de Buitenzorg, vii. pp. 213-23, with a 

 sketch map. 



