6i8 



NA I UKh 



[i >u HJisKK 27i '923 



M. 



Letters to the Editi 



ttot hoi 

 d by ht 



inlendeii 



I use they 

 re a« it 



\-! 



(>l \ 



ill V- 



It ; . ; . . h 

 (lir.. !.■.! IM il 

 i)t t lir all ii! ,1 i;- 

 pat tri 11, a lul el 1 : 

 alio! : 

 and ■ 



of ai <HHi 1! ]. 1 i ' 1 r M 



contribute t<i ■ In ( 1; 



ilifhciilties tba 

 tills in c()iis( (' 



f analvsi 



\slal sti! 



tl nlll will 



l-rruii 1! 



all iiu|uii\- iiiii.st 111' 



t \u- relative posit Kills 



■ iil.-^ within t lie unit 1 ■! 



\s liii li t he\- exert on oik- 



i ; ; r -x ii'imt 1 \ I li t li<- < r\ st al 



' I- I )t t lir Il'i il.ci liic ,ire I ] not |( III-, 



iill\ , i\ii|il 111 '-11 tar as tliev 



r I'l 'HI 1 l;i^ PI 1111 1 III \ ir'A I il(■ 



<!ifti^n!ties tiesi rilu li 1>\- Mr. I ;,irker seem to h i~>e niiK. li 

 of t lieir iinportanee. 



'IJie syiiunctry ilett nninat ions oi the oMer ami 

 well-known methods are (mniilete wln'ii the er\st,d 

 has been assigned to one or oilier of tlie tliirt\- two 

 classes : and this ean tjeneralK' be done, lli(ui,i;h tlu're 

 often reina.ins Munr inu ertaintw < )ii the other hand, 

 the X ra\' methods determine t lu" form and si/e of 

 the nijit I'l j'.'tteiii and liie iininber of molrcides 

 win- : of 1I . It IS an ojh'ii < 



\s ha ' .\\\ aiiswei", \\ het her tin 



f.;i\ e al 1 li.ilK eurreel e\ ideiiee on this piuni : 

 \\hether, Inr example, llie\' mij^lif oxciinok snnie 

 dittertMice wliieh, repeated at some multiple of the 

 spacin.u determineil by the ra}s. ini])lieil a larger 

 unit of i^adti'rn. The difference iniijht ]>i' of such a 

 kind as to be inea]iable of deteelion lio\\r\ar .L;real 

 it nii,L;ht be, or it niiL;hl be missed niereh' on areount 

 of insnllieient nia,i;nitiide. So far, the e\ ideiice points 

 to the pdNsession h\ tlic X-ra\s of th(> power to 

 delet t an\ material difference : the\- ean, feir exam])le, 

 iral.r aiManiM ;hr ^ ' i ifcrence between two neiqhbour- 

 ; the diamond which are dne onlv 



1 rientation of tlieir aftachnients 



to their neighbours. JMoreox t-r, the X-ra\s L;i\e such 

 information as to the relatixe positions of the atom- 

 groujis mti) v\hirh tlie (iN^tal unit nia\- be dixided 

 that the (r\--tal ^aii imt onl\- be assii^ned to its proper 

 class, Init also to its proper type amoii^ the two 

 hundred and thirt\ possible types of structure, as 

 dehned either by the space :• ;> ' svmnietry move- 

 ments wliieli the crystal ]>i ■ h\ the arrange- 

 ment which the X-rays have loimd. There is onlv 

 one c.xcei>tion of importance. 'I'he X raxs cannot 

 in general determine whether a crxst.d has or lia> r ■ 

 a centre of symmelry ; aiul this sometimes Km 

 the alternatue as to \shethcr a er\stal belongs to d 

 certain t\pe of syninu'trx' in a lower class or to 

 anotlicr t_\pc in. a liigher cla.ss obtained from tlie 

 lower by adding the centre of symmeirw If it is 

 known from observation of form or otherwise whether 

 there is or is not a centre of sMnmcirv the ambiL;uity 

 disappears. 



The general arrangement (^f the molecules 1-aving 

 been found, any furth(>r observations of the form or 

 other pin ■ " [^crtics of the crystal have a \alue 

 "Ot bee. may put llie crystal up or down 



NO. 28 17, VOL. I 12] 



-•ay that licre is the m 

 preferred, that the mol 

 iganic crystal, naphthalene 

 ■torn group whirn mav wi' 



4 to the n;o 

 _;s, and onn tf> 

 M( iiie. is tiicre any 

 iie attachment of the • 

 1 \ liiogens - If there is, then there i> e, 

 thing which may l>e looked on as the i 

 there is a group which has the same C' 

 the free naphthalene molecule and \n 

 form such a molecule on the di.ssolution o. ..,, ,. > 

 thiiu.^h the shape might be slightly clianged. If i 't, 

 :! a I it might be said, as of the diamond, tlion e. • , 

 degree in this case, that the who! 

 le molecule. The jx)sition of the ; 

 would be an example of " co-ordination." The 

 nature of the h\ !r ■_! n attachments is obviously of 

 the highest impo and we may hope to learn 



more about it b' ''xperiment. Only in that 



liuht, hf 'crest in discussing the 



ipiestioi : the molecule in the 



naplithaleia W. H. Bk\e,<,. 



The Optical Speetruin <if Hafnium. 



I N our letter to of March 10, 1923, we gave 



a preliminary list 1 ist prominent lines between 



J500 and ,^500 A : 'ectrum of the new 



element hafniui! "oster and Hevesy 



SCI - , January .io, ceoruary 24, April 7, 1923). 



in we included only lines of intensity of 2| 



and iiiwit.-, ll^ing a scale of intensity from \ to 6. Prof. 

 Hevesy has now been able to supply us with a pure 

 hafnium preparation containing according to X-ray 

 analysis not more than about i per cent, zirconium, 

 and ill addition small traces of titanium, manganese, 

 and niobium. W ith this preparation we have photo- 

 graphed botii the arc ,111 1 the spark spectrum of 

 hafnium. As to the otnim, we have con- 



trolled all our old. r aents of the published 



lines, and also (c iver unpublished lines, and 



have added to t 'usiderable number of still 



weaker lines. 



The spark -pcctnii;; :i hitherto to have 



I )n ill iiii; r luuul, Bardet (Cow/)/f^ 



I 711. 1 )-p a short time ago pub- 



ii-:ic ! a i>t I'l imes belonging to the arc spectrum of 



hafnium m the region between 2300 and 3500 A. U. 



' • not stated and cannot be inferred whether 



are given in the international or in the Row- 



..; w -vale, it is sometimes difificult to decide whether 



a line in Ins table coincides with one of our lines or not. 



v-; lii hi- ic^,.-; \ve:i ih,. cvc.-iit 1, m of (j are stated as 



cient faible," " a 

 \ / - - preparation was 



not iiole, the tvvo spectra 



are irtiet does not find all 



the 1 m our hrst table, which (with 



one A all confirmed, and the relative 



intensity of ins lines are rather different from ours. 

 This seems to indicate that the hafnium spectrum can 



