1878.] HI TBarker. 



was, at your suggestion, so adjusted that one of its needle-points rested, on 

 the hydrogen line near G and the other on the line known as It. After the 

 last ray of sunlight had disappeared, I took a few seconds of the precious 

 time to observe the eclipse with the naked eye. The moon appeared in- 

 tensely black, surrounded by a pinkish halo, extending to about two fifths 

 of a lunar diameter from the limb, and occupying the entire circumfer- 

 ence. At two points this halo was expanded into radial streamers, one of 

 which had parallel sides with a deeply indented or swallow-tailed end, ex- 

 tending Avestward of the sun and apparently lying in the ecliptic ; the other 

 appeared single, was on the eastern edge, and was inclined twenty degrees 

 or more to the north of the ecliptic. The former of these streamers was 

 traced to a distance of about a lunar diameter and a half from the edge, the 

 latter to a somewhat less distance. No structure could be seen in the halo, 

 but in the streamers traces of parallel rays appeared to be present. The 

 amount of light emitted by the corona was a surprise to me. Preparations 

 had been made for using artificial light for reading the circles, but this was 

 found not to be at all necessary. The amount of light seemed to be nearly or 

 quite equal to that given by the moon when ten days old . No protuberances 

 were seen with the naked eye ; nor were any streamers observed, other 

 than those already described. A glance at the eclipsed sun was then taken 

 through the finder of the equatorial. The magnifying power being low, 

 the corona presented much the same appearance as to the naked eye ; but 

 the streamers showed much more distinct evidences of a radiated structure 

 and a pale rosy protuberance was observed on the south-western edge of 

 the dark disk. This was undoubtedly the same prominence which was 

 observed previous to totality. 



Turning my attention now to the spectroscope, upon the slit of which 

 the coronal image had already been brought by means of the finder, the 

 slit being placed radially, the first, glance through the instrument showed 

 me a bright, but an absolutely continuous spectrum. The region under 

 examination was of course that portion of the spectrum which had been 

 placed before totality between the needle-points of the micrometer. Totally 

 unprepared for so unexpected a result, I moved the observing telescope so 

 as to bring the green portion of the spectrum into the field, expecting cer- 

 tainly to see 1474 K, and by the appearance of this line to determine whether 

 my instrument was out of adjustment ; and if it were, to adjust it again. 

 But no bright line was there ; the green region appeared as continuous as 

 the blue. I then gradually closed the slit — which had been previously ad- 

 justed on the solar spectrum so that the line D appeared nebulous on its 

 edges — thinking that I might in this way improve the definition, but with 

 no better results ; no bright lines could be seen. To my great surprise, 

 however, when the slit was thus narrowed, the region which was then 

 under examination, that extending from b to G, appeared filled with dark 

 lines on the brighter background, these dark lines being readily recognized 

 from their general appearance as the solar lines of Fraunhofer. Still intent 

 on getting bright liues, I opened the slit again gradually, moved the observ- 

 ing telescope over the entire length of the spectrum from red to violet, re- 



