166 Suggestions for the Observation 



last complete contact and the last disappearance of beads 

 (or other irregularities in or about the cusps), should be de- 

 termined. 



The remarkable fact of a recurrence of cusps, and the pos- 

 sible explanation of it, should be attentively considered. 



In reference to the phenomena of the corona, and their 

 possible explanation, the observer is referred to Professor 

 Powell's papers in the Memoirs of the Astronomical Society y 

 vol. xvi., p. 301, " On Luminous Rings round Shadows," vol. 

 xviii., p. 69, " On Irradiation." 



Appendix No, II. 

 It is recommended to observers who may devote them- 

 selves especially to the phenomena of the rose-coloured pro- 

 minences : — 



1. Immediately before the total obscuration, to watch for 

 the appearance of the prominences on all parts of the sun's 

 limb, but particularly at the part just about to be eclipsed by 

 the moon's limb ; and, the moment that the sun's bright 

 edge is eclipsed, to watch in the direction of the moon's ad- 

 vance for any such prominence ; to note mentally its form, 

 and particularly the proportion of its height to its breadth 

 at the base, which may be afterwards recorded in a sketch ; 

 and to make quite sure whether or not the moon gradually 

 eclipses it. 



2. In like manner, and with the same view, to direct the 

 second scrutiny (immediately after the previous one) to the 

 diametrically opposite portion of the moon's limb, watching 

 for the summits of any prominences, and whether they en- 

 large as the total eclipse proceeds. 



3. In the third place, the observer having satisfied himself 

 on the two previous points, will carefully examine the moon's 

 limb all round, and may now record, on the previously pre- 

 pared circular diagram, the positions of any such prominences 

 round the moon's limb. Let this be done in the first instance 

 without regard to their form or size, but only with regard to 

 their distribution round the circle ; and let this be carefully 

 verified once at least. Let him note whether any kind of 

 arch of light connects two or more of the prominences. 



4. Let the dimensions and forms of the prominences be 



