390 



NA TURE 



[February 21, 1895 



point. This properly characterises the other occupants of the 

 eighth column ; to S'TOC extent in the iron group, and certainly 

 io the palladium and plitinum groups. 



(5) The introduction of a new element with the atomic 

 weight of 20 (not 21 or 22), will extend the range of certain 

 recurring numbers which appear ne.ir the beginning of the 

 series of at.imic wcrigh's. In the existing arrani;emen' of 

 the atomic weights it will be observed that between oxygen 

 l6 and fluorine 19 iheie are three units; between fluorine 

 19 and solium 23 there are four ; then the numbers run I, 3, 

 1.3, I, 3i. Adding argon at 20 the series becomes symmetrical 

 all the way fom oxygen to chlorine, as will be seen in the 

 diagram on the previous page. 



If, on the other hand, we were to adopt 40 as the atomic 

 weight of argon, we should meet with the following serious 

 difliculiies : — 



(1) There is no room for it. To place another figure just 

 before or just after calcium would disarrange the whole subse- 

 quent serits. 



(2) It would break the periodic law in regard to its meltmg 

 point. 



(3) It would break the law in regard to its atomic volume. 



(4) The inactive argon would be associated with metals of 

 the earths, the compounds of which are remarkably siable. 



■5) It W'uld iTing the atomic weight of three elements, 

 potassium 39, cilcium and argon each ahoul 399. wiihin one 

 unit. This never occurs elsewhere in Mendelerft"s lahle. 



.-Vgainst these consi.ieraiions there i-i the firci'ile argument 

 deduced from the ratio of the specific heats of argon. I will not 

 attempt to wei^jh the re-peclive meri's of these lines ot reason- 

 ing, especial y in the absence of the details of the experiments 

 on the velocity of sound, and until wc have some kmwled^e of 

 the compounds of argon. Trustworthy conclusions will not be 

 possible till this far her information is obtained. It is not a 

 question of physics versus chemistry, for the true theory of its 

 place am<>n,> the elements must be able to ciordinate all the 

 facts upon which both the chemist and the physicist rely. 



J. H. Gladstone. 



London, February 8. 



The Aurora of November 23, 1894. 



Per.mit me to c;ill attention to a significant fact disclosed by 

 a scrutiny of the observations of the aurora of November 23, 

 1894, and having an important bearing in discussing the auroral 

 dimensions, and which appears to have escaped notice (see 

 Nature, January to). I refer to the invisibility of the objects 

 at Dingwall, to the observers at Tynron, Dumfriesshire. 

 Extracts from the synchronous accounts in Nature, November 

 29, p. 107, and January 10, p. 246, will prove this slaleinent. 



Tynron, 7.30. — Luminous mist in the northein sky, strong 

 enough 10 cast shadows on the shining surface of the wet 

 ground. The mist m.ved from the horiz.m to the zenith, form- 

 ing a detached luminous belt in patches, disappearing at 8.30, 

 leaving only the light in the north. 



Din^.ratt, 7.30. — .Sky covered in all directions by a canopy 

 of streamers. At the same lime the arch Hisappc.ired, and 

 occasional streamers up to eight o'clock. It is not possible 

 that the arch seen at Dingwall could be the same as the one at 

 Tynron, because the former had vanished when ihe latter com- 

 menced to form, whilst there is a total absence of streamers, and 

 phenomenally brilliant mist not recognised at the other place. 

 Until mcihiKls of observation and analysis can be introduced that 

 will eliminate the errors of idenliftcalion, the solution is likely 

 to be indefinitely postponed. W. H. Wood. 



Kirmingham, Fcliruary 3. 



Making the necessary allowances for increased apparent 

 lumino>ilic< of bright streaks, or of layers of light in the atmo- 

 sphere, by the foreshortening effects of end on, and of edge- 

 presentations, the observation* at Dingwall and in Dumfries- 

 shire of the aurora of November 23 last, scarcely seem to recount 

 very much which wis not at the two places, at least partially, a 

 fairly comparable anrl nearly conlemporancous description of 

 the same phenomena. The first-formed light-band of the glow 

 was very strong at Dingwall from ihc east to west, a litilc 

 southward from the zenith, until 7.30 p.m., when wilh Ihe 

 usual drift of such displays lo soui lizards, it became less 

 prominent theie than the approaching canopy of streamers 



NO. I 32 I, VOL. 51] 



which supervened, drifting up from the north-eastern sky in rear 

 o( it. Hut it assumed at ihe same time increasing pri'minency 

 in Dumfiiesshire (150 miles south of Dingwall), where between 

 7.30 p.m. and S p.m., it passed overhead in ihe shape of de- 

 tached patches of light, a form which the belt was also seen lo 

 as.sume and to break up into, in a slow extinction st.ige at 

 Slough, in that interval. The display of streamers rising from a 

 large tract of light-mist approaching Dingwall from the north- 

 ward and increasing constantly in lustre at 7.30 p.m., did not 

 extend far south of Dingwall before it faded out soon after- 

 wards ; and being (as it was seen at .Slough) a dense local dis- 

 charge of them, its corona of brightly-foreshortened beams over- 

 head would naturally be a very impressive sight at Dingwall, 

 although from a positi n 150 miles oistant in Dumfriesshire, the 

 broad-.iHe aspect of the short outbreak, seen from afar, instead of 

 from underneath, would only have the appearance of a sheaf of 

 coloured light projected up from the usual flat streamer-base, 

 neither very «ide nor extraordinarily lofiy, but of the m.issive 

 berg-like form, which was its description at Slough, not un- 

 frequemly noticeable in rather strong auroras ; and it may even 

 have been quite easily hidden from view entirely at Tynron, 

 although a clear horizon in the north near Slough allowed its 

 observation there, by trees or by other obstructions in the 

 landscape. 



If Mr. Wood can happily devise a practical means of per- 

 fectly recording the times and descriptions of all the rapidly 

 changing features of an aurora, and the shifting variations of its 

 misty light-glows, he could no doubt achieve results from ob- 

 servations which would he no less a bent lit to astronomers and 

 terrestrial magneticians, than an exact continuous registration 

 of cloud phenomena would lie of welcome interest to meteoro- 

 logists and terrestrial clectiicians ; but it needs no great famili- 

 arity wilh auroral exhibitions to be quite certainly assured that 

 results of records even so elaborate as those might be, would 

 never be found to confute, but only to confirm the generally 

 accepted view of the really cosmical heights and dimensions of 

 all truly auroral lights and corruscalions ; of all such lights, 

 that is to say, as show in their spectra the yellowish auroral 

 line, or some of the other well-recognised spectral indices of 

 the aurora. A. S. Herschel. 



The American Association. 



At the Brooklyn meeting of the .\merican Association for 

 the .-Xdvaiicement of Science last summer, it was decided lo 

 meet this year at San Francisco, provided reasonable rates of 

 fare could be secured from the trans-continental routes, as it 

 was supposed could be done. Prof. Joseph Le Conte for threfr 

 consecutive years had crossed the coniinent, laden with earnest 

 and cordial invitations from the Universities and .Scientific 

 .Societies of California, and the Common Council of San Fran- 

 cisco, to hold the meeting for 1895 in that city. The short- 

 sighted policy of the railroads, however, refused to grant any 

 concessions ; and it has at List been deciiled to meet at Spring- 

 field, Mass., August 29— September 4. The meetings of 

 affiliated societies will begin on Monday, August 26 — rather 

 later than the usual time of meeting. 



Springfield is a small city, located in Ihe heart of New 

 England. It is the scat of the principal arsenal of the United 

 Slates ; and, while not a University city itself, it is within two 

 or three hours' ri<le of nearly, or quite, a score of institutions of 

 learning of the highest grade, including the two oldest and 

 most powerful Universities in America, Yale and Harvard. 

 This will be the second meeting at Springfield, the first having 

 been held in 1859. 



The Association is incorporated by the Stale of Massa- 

 chusetts, and its office and museum are at Salem in that Slate J 

 but no meeting has been held in New Kngland since ihat at 

 Bo.sion in 1880, the most brilliant in all the history of the 

 .Association. The return to New England after this longest 

 absence, gives unusual interest to the approachirg meeting. 



W.\t. H. llALi-;. 



Brooklyn, New York, February 2. 



I 



Earthquake in Norway. 



AiioiT midnight on the 4lh and 51I1 of this month, a fairly 

 strong earthquake occurred in the snutticrn part of Norssay. The 

 greatest di-tuibance was fell in the environs of the town of 

 Aalesiind, upon the west coast (about 60" 30' lat. N.). From then 



