35S 



NATURE 



XAiig. 29, 1872 



the present year (1S72), the number of cyclones is already (28th 

 June) greaU-r than in any year since 1 861. 



What has now been said is not only liorne out by the records 

 of the Meteorological Society, which give detailed accounts of 

 the hurricanes, but also, I have little doubt, by the books of the 

 docks and marine establislmients. 



Especially in 1847-4S, and again in 1S60-63, the harbour of 

 Port Louis was at limes crowded with disabled ships ; whereas 

 in the years 1855-57 and 1S6G-68 there were very few. 



It win be seen that these years correspond pretty closely with 

 the maxima and minima epochs of sun-spots. 



For the present I wish merely to call attention to the subject, 

 in order that the connection which I think exists between sun- 

 spot frequency and cyclone frequency may be either verified or 

 refuted by jiast or future observation. 



It appears to me that there is more than a mere coincidence as 

 to time. Tliere are three maxima and two minima epochs of 

 cyclone frequency, corresponding nearly, if not entirely, with 

 similar sun-spot epochs. 



To examine the matter fully, it would be necessary not only to 

 know the number of cyclones in each year, but also the extent 

 and duration of each and tlie force of the wind. If we could 

 thus get an expression for the annual amount of cyclonic energy, 

 and could show that it varied directly as the amount of sun-spots, 

 a connection would be established. One violent hurricane, which 

 lasted ten days and passed over thousands of miles, might have 

 more value than half a dozen smaller and short lived ones. 

 However, having traced a large number of the cyclones in ques- 

 tion, I have no doubt that the years of greatest cyclone frequency 

 were generally, if not always, the years of greatest cyclone 

 energy ; and that the number of cyclones in a year is a fair ex- 

 pression of the cyclonic activity for that year. 



Now, taking the maxima and minima epochs of the sun-spot 

 period, and one year on each side of them, and comparing the 

 number of cyclones in these three-year periods, we get the follow- 

 ing results : — 



Number of Cyclones Total number 



Years. in each Ye.ir. of Cyclones. 



( 1847 4 ) 



Max. 184S 6 ... 15 



( 1849 5 \ 



( 1855 4 ) 



Min. \ 1S56 I ^ ... 8 



( 1S57 3 \ 



1859 5 



i860 8 



1S61 8 



I 1866 5 



Min. \ 1867 2 



( 1868 2 



( 1S70 3 , 



Max. 1 1871 4 \ ... 14 



( 1^7^ 7 ) 



Taking two years on each side of the so!ar-spot epochs, we 

 got :— 



Ve:irs. Number of Cyclones. Tot.il number. 



/ 1854 3 > 



1855 4 



Min. \ 1856 IV ... 15 



■857 3 \ 



^ 185S 4 ' 



, 1858 4 V 



1859 s 



Max. i860 8 ,. 32 



/ 1S61 S \ 



^ 1S62 7 ' 



I 1865 3 \ 



\ 1866 5 / 



Min. ■ 1867 2 \ ... 15 



/ 1868 2 \ 



^ 1S69 3 ■' 



Assuming that we have got a close approximation to the actual 

 number of cyclones, and that the numbers fairly represent 

 cyclonic energy, it is diflicuU to avoid the conclusion that the 

 above tables point to a definite law ; and that meteorology, mag- 

 netism, and solar physics are closely connected. For what Imlds 



good with regard to a large tract of the Indian Ocean probably 

 holds good with regard 'o other portions of the earth's surface. 



Is it not probable, also, that if there is such a connection as is 

 here suggested between the sun-spots, or sun-cyclones (as they 

 have sometimes been called), and earth-cyclones, there is a simi- 

 lar connection between the sun-spots and cyclones in the other 

 planets ? 



0)1 till- .'iptr/nim of Hydrogen, by Arthur Schuster, student of 

 Owens College. 



In a paper communicated to the Royal Society, I have 

 shown that nitrogen has only one spectrum ; the band spec- 

 trum usually obtained at low pressures being due to oxides of 

 nitrogen. I have since subjected hydrogen to a similar investi- 

 gation, and although my experiments may not seem to give an 

 absolute proof of the correctness of the opinion advocated 

 chiefly by Angstrom, they show clearly, I think, to what causes 

 we have to ascrilie the different results obtained by difl'erent ob- 

 servers. The changes through which the spectrum of hydrogen 

 passes when the pressure is giadually diminished are, according 

 to WUllner,* as follows : the spark of the small induction cod 

 begins to pass under a pressure of 135 mm. ; the colour of the 

 spark is white, but only under a pressure of 100 mm. it becomes 

 sufficiently intense to be examined. The spectrum is a continu- 

 ous one with shaded bands. This spectrum gradually increases 

 in brilliancy down to a pressure of 30 mm. If the pressure is 

 diminished still more the three lines llo, H;3, II 7, rise more 

 and more from the contiinious ground and the band spectrum 

 gradually disappears. Under 3 mm. pressure only these three 

 lines are seen, but if we continue to exhaust the tube, the con- 

 tinuous spectrum reappears in the green. Finally at the lowest 

 pressure under which the spark passes a third spectrum of bril- 

 liant lines is seen. A short historical sketch will show in how 

 far the above .statement is in harmony with the results of 

 Plucker, and will justify at the same time the course of experi- 

 menting which I have pursued. 



The spectrum of hydrogen obtained by the passage of the in- 

 duction current through vacuum tubes was first examined by 

 Plucker in the year 1S58, and described by him as follows •.\ — 



Hydrogen gave a comparati\ely simple spectrum, in which 

 five bands of about the same width are most prominent ; a violet 

 band A at the limit of the spectrum. Three bands in the green, 

 B forms the limit towards the violet, and D towards the yellow. 

 C is double the distance from D than from B. The fifth band 

 E is in the yellow. The most intense of the three green bands is 

 D, the colour of which is already yellowish. Next in intensity 

 is C, and then comes B. The red is very prominent and extends 

 over a wide space ; a well-defined broad black line situated near 

 its external limit ; another broad black line separates it from 

 the yellow band E. 



The description of this spectrum agrees in its general features 

 with the continuous spectrum described by WiiUner, and it is 

 therefore important that already in August of the same year 

 PliickerJ publishes a second paper, in which he says that the 

 above experiments had only been preliminary ones, and that the 

 hydrogen had not been pure. Pure hydrogen, he says, shows 

 only three lines, a red one, a greenish blue one, and a violet one 

 which is less bright. Plucker makes repeatedly the remark that 

 the purer the hydrogen the more concentrated is its light § on 

 the above three lines, and he cypresses the opinion that the bands 

 seen in impure hydrogen are due to air. Later on, li when he 

 was looking out for different spectra of the same gas, he found 

 that this spectrum cannot be due to air, and must therefore be a 

 second spectrum of hydrogen. It must, however, be remarked, 

 that Plucker never found this spectrum so well developed as 

 WiiUner, and this cannot be due to the fact that he did not exa- 

 mine it under the same circumstances, for he describes ^I the 

 change which the spectrum of hydrogen gas undergoes when the 

 pressure is increased, and only mentions the expansion of the 

 lines. Although the Leyden jar was introduced in these experi- 

 ments, and thesecondspectrum would not therefore haveappeared, 

 it is not to be expected that such a widely different spectrum 

 which changes entirely the colour of the spark should have 

 escaped the notice of such an observer as Plucker wdien he was 

 looking out for different spectra. The mostj important recent 

 contribution to the history of this spectrum has been made 



* Fortchrift der nieiierrheinischen Gesellschafc zur Feier Jes 50 jahrigcn 

 Subilaums der Univ. Bonn. 



t Pogg. Ann. civ. p 113. J U'id. cv. p. 76. 



§ Ibid. cvii. p. t:o7 and p. 518. 



11 Ph. Trans 1863, vol civ, H Phil Mag. xlii. p. 395. 



