PHOTOGRAPHS OF THE CORONA. 121 



sarily the type which is associated with a comet of extremely 

 small perihelion distance. The comets of 1843 and 1880 are 

 conspicuous examples. But this would leave unexplained the 

 synclinal structure from whence the long rays proceed, and of 

 which they form the continuation. The long-exposure photo- 

 graphs, which we are discussing, necessarily give us absolutely 

 no information as to the way in which these synclinal structures 

 are built up. For that we have to turn again to photographs 

 of larger scale and short exposure. 



These show us that over the principal prominences, and at 

 some little distance, an arch of coronal matter is formed. This 

 is succeeded by a larger arch outside, and so on for a succession, 

 the outer arches being less definite and complete than the inner 

 ones. Outside all we find the curves defining the boundaries of 

 the synclinal group. 



The impression, which the study of these forms has produced 

 upon us, is that the prominences represent centres of strong 

 eruptive action, and that in consequence of such action coronal 

 matter is driven upward from the sun over a very wide area in 

 dome-like forms. These as they rise would seem to meet with 

 some resistance for a time, a resistance most effective near the 

 circumference of the eruptive area, which averages about 40 of 

 a solar great circle in diameter. But near the centre of this 

 area the eruptive force may be strong enough to overcome the 

 resistance whatever may occasion it and therefore from the 

 apex of the synclinal structure we find the coronal matter 

 driven outward in a straight line, which probably indicates an 

 immense velocity. It must be noted that this eruptive action 

 is not always radial. One of the long rays in 1898 was tan- 

 gential, and another was oblique. Now a ray which is truly 

 radial must appear to be such whatever its position, whilst the 

 effect of perspective^ may easily cause lines to appear radial 

 which are not nearly so. 



In brief, then, it seems to us that the corona may be con- 

 sidered as threefold in character. First there is the lower 

 corona, pretty evenly distributed round the sun, and showing 

 here and there forms which recall on a somewhat larger scale 

 and with much diminished luminosity the typical forms of the 

 prominences. Next and it is this section of the corona on 

 which our present photographs have thrown light we have the 

 synclinal groups rising over disturbed regions, and at least in 

 the best developed instances shooting out at the apex into long 

 rod-like rays. Thirdly, we have a great number of little solar 

 comets scattered indifferently round the sun, but most easily 

 seen near the poles and at the sun-spot minimum. All these 

 three together, projected one upon the other on the apparent 

 vault of the sky, make up for us that complicated structure 

 which we call the corona. 



