io6 



NATURE 



[May 30, 1S95 



of its excursions in starch <if pond life, the neighbourhood visited 



\k\x\s, Tolteridge and Mill Hill. Mr. \V. Burton obtained 



- "'! phials of the water for examination, and the first 



«ater turned out into the trough contained a minute 



Mr. Burton kindly brought tome, when I immediately 



i(ientiried it as the Polyiicma iiataiis (Lubbock, Trans. Linn. 

 So,:, vol. xxiv. 1864, p. 135, plate 23). 



.\s this capture was, for the fourth time, the result of chance, 



Mr. Burton and I set out (May 6) to search for more s|)ecimens. 



' . r nets in and carefully examining the contents 



r<, my patience was at last rewarded by seeing a 



., stniggling to free its wings from the mass of 



minute vegetation gathere<l in the dipping net. Aher a few hours 

 more search, I found four males, which, together with the female, 

 I transferred to an observation tank, where all soon disportetl 

 themselves in the liveliest manner, swimming, or rather flying, 

 under water for over four days, during which period they did 

 no., to my knowledge, once leave the water. I have since obtained 

 others, which are under close observation, and in course of time 

 I hope to trace out their life-history. 



I'erhaf s, owing to the microscopic dimensions of many of the 

 MymariJu (Haliday), very few entomologists have paid any 

 attention to this most interesting and fascinating family of 

 beautiiiil " Fairy Flies,"' to whose industry we are no doubt 

 largely indebted for our freedom from " blights" of many kinds. 

 They ar.-, indeed, mere specks, scarcely visible to the eyes of 

 ordinary folk, and yet they have their place in nature. 



I am incline<l to think that when the ty|X- collection of the 

 MymariJu, made by the late Mr. Haliday, has hieen thoroughly 

 cxatniced, this name Polynetna natans will have to give [ilace, so 

 far as the genus is concerned. I ho|>e that before very long we 

 shall have figures of all the genera in this most interesting group. 



Fred. Knock. 



Halley's Chart. 



1 HAVE been much interested with the letter of Dr. L. .\. Bauer 

 in your la-st numlwr, as I happen to possess a map, or chart, 

 iKiund up with a number of Dutch, German, and French maps 

 of the end of the sevtnieenth and the first years of the eighteenth 

 centuries. The latest map with a date is 1704. This English 

 map is evidently the sime as 974 (4) mentioned by Dr. Bauer. 

 It is entitled " A new ami correct chart showing the Variations 

 of the Compass in the Western and Southern Oceans, as ob- 

 seivcil in y« year 1 700, by his Ma"'=^ command by Edm. Halley." 

 The dedication reads .is follows, in Latin : " .Majestati semjier 

 August.c Clulielmi HI. D.G. Magn.e Britannia.- P'ra. iV Hib. 

 Kegis Invictissimi. Tabula \vxc Hydrographica \"arialionum 

 Magncticarum Index. Devotissime Consecratur a Subdilo 

 Humillimo Edm. Halley.'' At one side of the map is the fol- 

 lowing : " The curve lines which are drawn over the seas in 

 this chart do show at one view all the places where the variation 

 of the compo-ss is the same : The numtjers to them show how- 

 many degrees the needle declines either E.Tstwards or Westwards 

 from the true North : and the double line passing near Bermudas 

 and the Caf)c de Virde isles, is that where the needle stands true 

 without variation." 



The chart is in excellent condition, but has no name or 

 printer on it. The only indication is " \. Harris, Sculp." The 

 course of a vessel going from and reluming to England is clearly 

 niarke<l. Tnos. Ward. 



Northwich, May 27. 



".\ ////•. I.I.M: .N/V:( /A'.-; nh I//E 

 ELE.MEMT.S. 



I THINK Lccoq de Boisbauclran was the first who 

 called attention to the fact that the line spectra of 

 the elements are by no means so irregular as they seem 

 10 be at first sight. He discovered the similarity in the 

 spectra of the alkalies and alkaline earths, and pointed 

 out how the lines in the spectra of these two families seem 

 10 be shifted towards the less refrangible side with in- 

 creasing atomic weight. .Mascart, in 1869, found two 

 strong triplets of lines in the ultra-violet spectrum of 

 magnesium, similar to the strong green triplet so pro- 

 minent in the solar spectrum. He says: "II semble 

 difficile que la rcproduclion d'un parcil ph<!nominc soil 



NO. 1335, VOL. 52] 



un cfTet du hasard ; n'est-il pas plus naturcl d'admcttrc 

 que CCS groupcs des raies seiiiblables sent des hamioniqucs 

 qui ticnnent i la constitution moleculaire du gaz luini- 

 neux ? II faudra sans doute un grand noinbrc d'obscrva- 

 tions analogues pour dccouvrir la loi qui regit ces 

 hannoniques." But the wave-lengths corresponding to 

 these rays were then not accurately known, and so the 

 I most interesting feature concerning the oscillation fre- 

 I quencics, or the number of waves which pass any fixed 

 point in unit of time, remained unnoticed. It was later 

 on shown by Hartley, that the differences between the- 

 wave-numbers of the three lines seem to be the same for 

 all the triplets. This constant difference of wave-numbers 

 repeated in a number of doublets, of triplets, and of more 

 complicated groups of lines, has now been observed in the 

 spectra of many elements. There are repetitions of 

 doublets in the spectra of sodium, potassium, rubidium, 

 caesium, copper, silver, aluminium, iridium, thallium ; of 

 triplets in the spectra of magnesium, calcium, strontium, 

 zinc, cadmium, mercury, manganese, and of more compli- 

 cated groups of lines in the spectra of tin, lead, arsenic, 

 antimony, bismuth. In all these cases the differences 

 seem to be absolutely constant. For, notwithstanding the 

 great accuracy with which Rowland has taught us to 

 determine the wave-lengths, the law holds good. As an 

 example, 1 give the list of doublets in the spectrum of 

 thallium, according to Prof Kayscr's and my determin- 

 ations. The number of waves passing a fixed point in 

 unit of time, is ec|ual to the distance the light travels in 

 unit of time divided by the wave-length. If we measure 

 the wave-lengths in vacuo, the distance the light travels 

 is the same for all rays. We may then choose as unit of 

 time, the time that light requires to travel one centimetre, 

 so that the wave-number is simply equal to i X, X being 

 the wave-length in vacuo, measured m centimetres. In 

 this manner, we get rid of the necessity of settling the 

 velocity of light, which as yet has not been measured with 

 anything like the accuracy with which the wave-lengths 

 are known. 



I 'x 

 iS6S4-2"l 

 26476-6/ 

 28324-1 \ 

 36117-1 ( 

 30952-1 1 

 38744-8! 

 335694 \ 

 41365'^ 

 3421771 

 42010-2 ( 

 34526-2 ( 

 42321-4 ( 

 353721 I 

 43164-71 

 36879-2 I 

 4467 1 -0 ( 



375030 > 

 45293-8 I 

 38305-01 

 46096 -8 ( 



46452-4 ( 



39'57o( 

 469473J 



The mean of the twelve differences, assuming their 

 weights to be inversely proportional to the fquare of the 

 cstim.'iled limit of error, is 7792'5. When the wave- 

 lengths are not reduced to vacuo, the differences are also 

 very nearly constant, because the reduction alters them 

 all nearly by the same amount. Hut it was a source of 

 satisfaction to me, that the reduction brought all the devi- 

 ations from the mean value well within the limits of error, 

 where.is without the reduction the second diffeieni e had 

 bccTi just beyond the limit. These tweb e doublets do not 

 comprise half the number of wave-lengths that have been 



