Ill 



before the eye, the impression of the object as a whole would be 

 retained upon the retina. It was, however, found both by Hug- 

 gins and ZoUner that it is only necessary to open the slit, placed 

 in a tangential position to the sun's limb, to the width required, 

 and the whole form of the prominence can be observed. This 

 result is simply a natural extension of the principle explained 

 above of diluting the continuous spectrum due to reflection near 

 the limb, until it is rendered invisible, when the chromatic prom- 

 inence lines stand out clear and distinct. Now lines are only 

 images of the slit, therefore open the sht when the continuous 

 spectrum has been so diluted, until it contains the whole prom- 

 inence, and the prominence will paint its own spectroscopic 

 images. Mr. Huggins was the first to practically employ the 

 method. His announcement to the Eoyal Society is dated Feb- 

 ruary 16th, 1869, and the first prominence ever seen by man in 

 all its details on the uneclipsed sun had been observed by him 

 three days previously. The method was at once adopted by the 

 whole band of observers, and many most interesting and valuable 

 results have been secured in the domain of physical astronomy 

 by the labours of such men as Lockyer, Secchi, Young, Resphigi, 

 Zollner, and others. 



But one year's observation of solar flames by the open slit 

 method sufficed to enable observers to distinguish two classes of 

 these objects; and the broad division of "cloud-prominences" 

 and " flame-prominences " was announced by Lockyer in April 

 1870. These classes are sometimes called hydrogenous and met- 

 allic from their chief constituents, or again quiescent and eruptive, 

 the first class being quiescent not always in themselves but relat- 

 ively to the eruptive class. Various subdivisions have been 

 suggested by Secchi, but it is very difficult to draw any clear line 

 of demarcation between the various forms. However, such forms 

 as plumes, horns, and clouds, either attached to the chromosphere 

 as is generally the case, or floating in the solar atmosphere, are 

 easily recognizable. Then, too, there are flames composed of 

 filaments which cross and recross in all directions, and again 

 diffused masses which rise from the solar surface, broad at the 

 base and tapering to their summits. Eruptive prominences are 

 generally of the cyclonic or spiked kinds, or if composed of fila- 

 ments, the filaments are aiTanged vertically. There are also such 

 shapes as sheafs or bundles of flame ; and jets and fountains are 

 to be seen in the neighbourhood of spots. 



We have before remarked that these prominences of the first 

 class are only relatively quiet. Indeed so rapid at times are 

 their motions, that the distinction of the classes into hydrogen 

 and metallic prominences, founded on their constituents would 

 seem to be the best. The first kind of motion we may call 

 attention to, is that caused by mighty solar winds or currents 



