248 



jVA TURE 



[January 12, 1899 



record, not only the above-mentioned 43-minute oscillations, but 

 also a series of much more rapid oscillations having a period of 

 about 35 seconds (the latter figure being obtained by a stop- 

 watch). The width of the basin is, at the point of observation, 

 2030 feet, and its mean depth about 100 feel (the tidal range is 

 less than 2 feet). Calculation by the above formula gives, for 

 first partial vibrations, a period of 35-8 seconds, which is very 

 close to the 35 seconds observed ; but it is to be noted that for 

 a basin whose depth is not small compared with its width, the 

 above formula is somewhat in error, and a more correct formula 

 (involving a hyperbolical cotangent) gives 37-5 seconds. In any 

 case the agreement is striking, and the two instances given 

 are almost conclusive as to the explanation advanced. Simul- 

 taneous observations on opposite sides of such a basin would be 

 quite conclusive, but these I have not yet had an opportunity 

 to make. 



(4) The following notes on the other cases referred to in (l) 

 may be of interest in connection with observations referred to by 

 Capt. Thomson, i^uaco is only about twenty miles further up the 

 Bay of Tundy than St. John, and yet, while expecting to find 

 there a period similar to that at St. John, I found a period 

 of I2i minutes (from four separate records). This was at first 

 puzzling, but, later, an examination of the chart showed the 

 existence of a dangerous ledge (the <)uaco Ledges) coming 

 nearly to the surface, at a distance of eight or ten miles off 

 shore, and forming, with a headland above Quaco and another 

 below, an irregular basin, the dimensions of which no doubt 

 determine the period of oscillation. A quite similar explanation 

 applies to the oscillations at Halifax, for there a succes.sion of 

 banks (the Emerald, Sable Island, Le Have and Roseway) 

 form, with the Nova Scotia coast, a large-sized bay of irregular 

 shape. 



(5) As to the external impulse that starts the oscillations;, 

 there is much uncertainty. Marked oscillations at St. John are 

 frequently accompanied by barometric disturbances, but not 

 alway.s. My own observations make me believe that the oscil- 

 lations and heavy ground-swell usually coexist. A notable case 

 (for which I have to thank Mr. S. W. Kain, of St. John) oc- 

 curred on September 18, 1898, when the heaviest ground-swell 

 in several months was accompanied by marked periodic oscil- 

 lations recorded on the Kelvin gauge. On the whole, I believe 

 that the disturbance of equilibrium is due either to abrupt local 

 variations of atmospheric pressure, or to the transmitted effect of 

 a distant hurricane. 



Those who are interested in this subject will find fuller de- 

 tails of the cases here discussed numerically (and also a short 

 bibliography of the subject) in a paper by myself in the American 

 fournal of Sdence (vol. iii., 1897). The nature of the oscil- 

 lations in the Gulf of St. Lawrence are shown on curves that 

 illustrate a tidal report, by Mr.W. Bel! Dawson, in the last volume 

 of Transactions oi the Royal Society of Canada (see N.\TURE, 

 vol. Iviii. p. 260). A. W11.MER Dui'K. 



Purdue University, Lafayette, Ind., U.S.A. 



GENERAL LAW OF THE PHENOMENA OF 

 MAGNETIC PERTURBATIONS OF SPEC- 

 TRAL LINES. 



TN the Philosophical Magazine for April 1S98, I 

 ■*■ pointed out that the resolution into triplets, cJcc, 

 which the spectral lines suffer in a strong magnetic 

 field, did not appear to follow any obvious general law, 

 but appeared to be some complex function of the wave- 

 length. To this was added the following remark : — " It 

 is possible, however, that the lines of any one substance 

 may be thrown into groups for each of which fiX varies 

 as X-, and each of these groups might be produced by 

 the motion of a single ion. The number of such groups 

 in a given spectrum would then determine the number 

 of different kinds of ions in the atom or molecule. 



" Homologous relations may also exist between the 

 groups in different spectra, but all this still remains for 

 complete investigation." 



.'Xlthough this investigation is still far from complete, 

 yet the measurements so far made uniformly go to show 

 that the foregoing expectation is about to prove true, 

 and that <■ /// or 8 A X- is the same for the corresponding 



NO. 1524, VOL. 59J 



lines of the natural groups in the same spectrum, and, 

 further, that this quantity remains the same for corre- 

 sponding lines or groups in the homologous spectra of 

 different substances. 



Not only is the magnitude of the magnetic effect 

 governed by the foregoing law, but the character also of 

 the effect is the same for the corresponding lines ; and 

 this is very interesting, as it shows that the correspond- 

 ing lines probably arise from the same origin. The 

 theory is consequently verified by the facts when the 

 spectral lines are considered in groups corresponding to 

 the molecular events which produce them. 



Thomas Prkston. 



COAST-TELEGRAPHS AND SPACE- 

 TELEGRAPHY. 

 T^HE year 1898 was an important period in the history 

 ^ of space-telegraphy, it was the period in which the 

 possibility of being able to signal across wide stretches of 

 open sea, with certainty in all weathers and at high 

 speeds, became first generally recognised as practicable. 

 Within the year the final report of the Royal Commission 

 on the question of Coast-Telegraphs, published late in 

 1897, came into our hands ; and the last few months of 

 the year witnessed a truce to the war of " wireless- 

 telegraphy." .-X wave of good feeling has now united the 

 opponents into something like coherence, and the honours 

 have been divided with universal approval. The result 

 is that for the future Italy takes prominence, England 

 eminence, while Russia, Germany and P" ranee share the 

 luxury of many grievances. 



The close of the year is very appropriately characterised 

 by three papers, respectively communicated by Dr. 

 Lodge,' Mr. W. H. Preece,- and Mr. S. Evershed,' to the 

 Institution of Electrical Engineers, all emphasising the 

 merits of one and the same system of space-telegraphy. 

 The authors themselves were more or less unanimous as 

 to the course further experiment should take, but the dis- 

 cusssion that followed the reading of these papers showed 

 a certain lack of directness ; many of the speakers were 

 carried away by side issues, and a great deal of time was 

 occupied with ill-considered suggestions and old matter. 

 While fully recognising the value of open discussion, and 

 of hints thrown out at random on subsidiary matters, the 

 present writer thinks it may be useful here to indicate the 

 limits to which the problem may be narrowed down, and 

 to point out the very serious work that is now calling for 

 the aid of space-telegraphy. 



It is very generally admitted that space-telegraphy will 

 replace metallic-circuit systems only under conditions 

 where metallic circuits are impracticable. The fact that 

 metallic circuits have been laid over the .\ndes, may be 

 taken as proof that there are remarkably few land-areas 

 that cannot be spanned by wires. For communication 

 between fi.xed points on rough coasts, a wire suitably 

 protected is still the right and the best thing, as is 

 evidenced by the cable' laid in 1890 between I'ollagill IJay 

 on the north-west coast of Ireland, and Portdown Bay, 

 Tory Island, and thence by duplicate underground cables 

 to the lighthouse on the north side of the island. The 

 great advantage of a metallic-circuit system is the conse- 

 quent privacy of the inessages, the simplicity of the 

 apparatus, the speed of transmission, and the possibility 

 the system offers for working by telephone, and in other 

 ways avoiding the expense of skilled operators. .Space- 

 telegraphy is at present limited to comparatively short 

 distances, and its usefulness is confined to spanning 

 estuaries, skirting sea-boards, and for such purposes as 



1 " Improvements in Macnctic Space-Telcgr.iphy." 

 -" .Klhcric Telegraphy. 

 3 ''Telegraphy by Magnetic Induction." 



* See an impurtant paper by Mr. H. Hcnest, " Coasl-Tclegr.tph Com- 

 munication," read berorc the Balloon Society, March 18, 1693). 



