258 



NATURE 



[July i6, 1903 



that the " polar " and " equatorial " coronas are always 

 followed by " intermediate " types, the order being polar, 

 intermediate, equatorial, intermediate, polar, &c.^ 



The " intermediate " type may sometimes approach m 

 form a " polar " or an " equatorial " type, accordmg as 

 the epoch of the occurrence of the eclipse occurs nearer 

 or further from the epochs of occurrence of polar promin- 

 ences. 



Further, the " intermediate " type precedmg a polar 

 type will differ to some extent from one immediately follow- 

 ing a " polar " type, because the latitudes of the centres 

 of prominence action in each case are different, as can be 

 seen from the accompanying figure. 



Two coronas which have not yet been classified are those 



of 1865 and 1885. The former of these is of a type between 



the "intermediate" and "equatorial," while the latter 



falls between the "polar" and "intermediate" groups. 



These have been inserted 



on the dividing lines in 



Fig. I. In classifying the 



coronas into the above 



groups, I have been 



greatly assisted by Mr. 



W. H. Wesley, to whom 



I wish to express my best 



thanks. 



^he first natural and 

 crucial test to apply, in 

 order to determine whether 

 there was a connection be- 

 tween prominences and 

 the different forms of the 

 corona, was to inquire 

 whether the coronal 

 streamers were situated 

 nearest the solar poles, at 

 those times when the 

 prominences were most 

 numerous in those regions. 

 The comparison for this 

 test showed that the only 

 five " polar " coronas re- 

 corded since the year 1869, 

 when prominence observ- 

 ations were compienced, 

 occurred at those epochs 

 when the prominences 

 attained their highest lati- 

 tudes. 



This satisfactory result 

 indicated a very probable 

 cause and effect between 

 prominences and the 

 coronal streamers, for the 

 region considered was 

 quite outside the zone of 

 the spots, and therefore 

 independent of them. 



It was next found that 

 the other two types of 

 coronas were closely 

 associated with the number 

 and latitudes of the centres 

 of prominence action. 

 Thus the " equatorial " 

 type only occurred when there was one definite centre of 

 prominence action in each hemisphere, while the " inter- 

 mediate " type has been recorded at those times when two 

 centres of action in each hemisphere were in progress, neither 

 of which were in very high latitudes. 



The accompanying illustration (Fig. i) shows the re- 

 lationship between the sun-spot curve for both hemispheres 

 together, the latitudes of the centres of action of the solar 

 prominences for each hemisphere, explained in detail in a 

 previous number of this Journal (Nature, vol. Ixvii. p. 569, 

 April 16), and the times of occurrence of all the eclipses 



1 It may be here remarked that the "intermediate" type between an 

 "equatorial" and " polar " type has only once (1869) been recorded during 

 the period here under investigation, and this is due to the absence of 

 observed eclipses during the two short available periods since thit date, 

 namely, 1879-81 and 1890-92. 



INTERMEDIATE 



EQUATORIAL 



Fig. 2. — Diagram showing by radial 

 lines the positions of the centres of 

 prominence action and their relation 

 to the chief features of individual 

 coronal streamers and to the general 

 form of the corona. The positions of 

 the sun-spot zones are also indicated. 



that have occurred since the year 1857. When two eclipses- 

 of a similar type occur in the same, or two successive years, 

 they have been inserted either one above the other or 

 obliquely, respectively, to avoid overcrowding. A curve is- 

 also drawn through the different types showing their re^ 

 lation to the sun-spot curve. 



Since the systematic prominence observations only com- 

 menced in the year 1872, the dotted portions of the curves 

 previous to that date are intended only to give a rough 

 idea of the variations as based on a general repetition of 

 the observations of 1872 to 1885. 



Fortunately for the present inquiry, Respighi made some 

 very valuable prominence observations during the years- 

 1870, 1871, and 1872, which are sufficiently numerous to 

 indicate the positions of the centres of prominence activity 

 for these years. These showed that during the years 1870- 

 and 1871 there were two well-marked prominence zones in 

 each hemisphere, and that the latitude of one of the zones- 

 was very high. The positions of these zones are indicated 

 in the accompanying figure by the small dots against these 

 years, and they agree well with the suggested curves repre- 

 senting the probable conditions as might have been ex- 

 pected from subsequent observations. 



The different types of corona are plotted in three different 

 horizons in the order "polar," "intermediate," and" 

 " equatorial," and the symbols adopted for each, namely, 

 small circles with 8 rays for the first, 4 rays for the second, 

 and 2 rays for the third, are inserted at the epochs of their 

 occurrence according to the general time scale for all the 

 curves. The continuous and broken vertical lines denote 

 the epochs of the sun-spot maxima and minima, as deter- 

 mined from a discussion of spots recorded on both hemi- 

 spheres of the sun. 



At the first glance it will be observed that the three types 

 of the corona, as seen from the curve drawn through them, 

 follow the sun-spot curve very closely, that is, that at about 

 the times of the maxima of sun-spots, the " polar " type 

 is present ; at the minima, the " equatorial " type : and at 

 the intervals between these, the " intermediate " type. 

 Although the sun-spot curve thus affords a means of pre- 

 dicting in a general manner the epochs about which any 

 of these types will occur, such a small restricted zone which* 

 the spots occupy excludes the idea of their presence being 

 responsible for such widely distributed coronal phenomena. 



The prominence curve, on the other hand, not only pro-^ 

 vides one with a more accurate method of forecast, but 

 such phenomena can account for the general changes of 

 position and form of the coronal streamers. 



By examining the prominence curves in relation to the 

 three different types of coronas from the year 1869, this- 

 connection is seen to be very close. Thus, during the 

 years 1870 and 187 1 there were two centres of prominence 

 action in each hemisphere, one of which was in high lati- 

 tudes, and the coronas for that period were of the " polar "" 

 type. From the year 1872 to 1877 there were two centres- 

 of prominence activity in each hemisphere, both in com- 

 paratively low latitudes, and the two eclipses during the 

 period, namely, 1874 and 1875, were of the " intermediate "' 

 type. The next eclipse, 1878, occurred when only one 

 centre of action was in existence, and the form of the 

 corona was of the " equatorial " type. 



As these centres of prominence action reached their 

 extreme polar limits (about +80°), and a new centre had' 

 in the meanwhile commenced in lower latitudes (±25°), the- 

 eclipses of 1882 and 1883 were of the " polar " type. 



The next two eclipses, of 1886 and 1887, which were 

 "intermediate," occurred when there were again two 

 centres of prominence action in each hemisphere, but none 

 near the poles. When the centres became single, as they 

 did in the years 1S89, 1890 and 189 1, the two coronas 

 observed in the year 1889 were of the " equatorial " type. 

 With the movement of these centres to high latitudes in 

 the years 1892, 1893, 1894, the eclipse of 1893 was of the- 

 " polar " type. 



The two eclipses of 1896 and 1898, which were " inter- 

 mediate " in type, occurred when there were two chief 

 centres of prominence action, while the two most recent 

 eclipses of 1900 and 1901 were good examples of the- 

 " equatorial " type, and were concurrent with only one 

 centre of prominence activity in each hemisphere. 



If the eclipses observed between 1856 and 1870 be com- 

 pared with the dotted prominence curves for the same- 



NO. 1759, VOL, 68] 



