154 



as are the teeth in a comb, when the latter is held vertically 

 with the back of the comb below. At the lower edge of a 

 typical curtain the streamers are all united to form a con- 

 tinuous ribbon of light at an even altitude above the surface 

 of the earth ; above they fade into invisibility at varying 

 altitudes, imperceptibly passing into outer darkness. In 

 cases where the display is fainter, only the brighter sections 

 of a curtain may be visible, and the relation of such patches 

 of light in the heavens only become manifest at intervals 

 when the excitation increases and the intervening portions 6i 

 the curtain become dimly outlined. Often all that is visible 

 is one or two isolated streamers which experience has taught 

 are, in most cases at least, the visible parts of an invisible 

 curtain, if one may so express oneself. Curtains, as they are 

 viewed, wax and wane — now dazzling bright and again 

 quickly fading to almost invisibility. The excitation travels 

 definitely from one end to the other, often in slow, pulsating 

 movements. In the case of intense displays, however, the 

 luminescence ripples in a kind of wave-motion through the 

 streamers from one end to the other. In addition to the 

 passage of luminescence along the length of a curtain, the 

 whole curtain usually shows a tendency to drift (body-travel) 

 towards some quarter of the heavens. Curtains seldom 

 appear singly, but more often multiple, usually as strictly 

 parallel forms, in number proportionate to the intensity of 

 the aurora. Very brilliant curtains usually reduce them- 

 selves withim a few minutes to a nebulous condition, in which 

 the streamers and all fine outline are lost in a broad incan- 

 descent (nebulous) band or arc. Nebulous hand*, arc*, and 

 'patches are commonly referred to throughout this journal. 



Brilliant aurorae below the horizon may cast a reflection 

 on and so illuminate the borders of visible clouds, producing 

 an auroral glow. Our observations show that at other times 

 there was the appearance of auroral excitation connected with 

 specific clouds. 



The term corona is not employed here, for in literature it 

 is, often at least, misapplied. The streamers of curtain rings, 

 when such are situated in the zenith, on account of perspec- 

 tive effect, appear to converge to a meeting-point above, and 

 so have often been referred to in literature as a special form 

 under the name of corona. Elsewhere in contributions on 

 the subject of the aurora polaris the term is employed to refer 

 to manifestations when an intense display reaches th& zenith. 



The term arch is, except perhaps in exceptional cases, a 

 misnomer, as extensive linear bands are of the nature of arcs 

 at a fairlv uniform height above the surface of the earth. 



