of the Total Eclipse of July 28, 1851. 167 



studied. For the former purpose reference should be made 

 to the two parallel threads in the eyepiece of the telescope. 

 For the latter, observe — Whether the prominences have hard 

 and permanent, or waving and ill-defined outlines. Whether 

 they are invariably broadest at the base, and have on the 

 whole a tapering shape. Whether they seem to stand erect, 

 or whether any or all of them are aslant, like teeth on th0 

 edge of a circular saw. Whether any of them taper inwards 

 next the dark limb of the moon : whether they appear isolated ; 

 and, if so, how the space between the red patch and the 

 moon's limb is occupied. Whether the prominences vary in 

 outline during the scrutiny. Whether any appear to grow up 

 or to diminish ; and, if so, whether such change is what the 

 moon's motion would naturally account for. 



5. Let the illumination of the prominence be studied. First, 

 as to general colour; by inspecting them with the undefended 

 eye, both with and without a telescope (without any dark 

 glass). Next, as to distribution of colour, select a well-de- 

 fined prominence and examine it all over repeatedly with a 

 considerable magnifying power, and observe if it appears 

 absolutely uniform in colour and brightness, or whether it 

 shews any marks of structure or shadow or variation of tint. 

 It seems very difficult to suggest any comparative experi- 

 ment for recording the brightness of the illumination of the 

 prominences. 



6. As the total phase goes ofi^, let the eye be fixed on one 

 or more of the prominences, and see whether they instantly 

 and totally vanish, or for how many seconds they can be kept 

 in view. 



It may be well to refer to M. Arago's narrative of what 

 was seen in 1842 and on former occasions, in the Annuaire du 

 Bureau des Longitudes for 1846, and to a paper by M. Faye 

 in the Comptes Bendus de rAcademie, 1850, Nov. 4. 



Appendix No. III. 



Allusion having been made to instruments for determining 

 the plane of polarization, it may be proper to give the fol- 

 lowing information : — 



Nicol's prism is described in the Edinburgh Philosophical 



