1865.] Prof. Phillips Physical Aspect of the Sun. 51 



the annexed diagram, not those which really occurred, as shown in the 

 diagram already referred to. 



If we regard the spot as marking an opening in a bright photosphere, 

 through which the dark body of the sun is seen within a terraced penumbra, 

 as represented in the diagram of Sir J. Herschel* (here copied), it is obvious 



that toward the edge of the sun, when the line of sight, E c, forms a very 

 small angle with the surface of the sun, one side of the supposed terrace b 

 will become invisible if the intervals of height between b and a be very con- 

 siderable, and in the same way a part or the whole of c will be lost. The 

 photosphere of the sun, then, if it exist, appears not to have the enormous 

 depth sometimes ascribed to it ; or if we suppose the spot, with its ring 

 and border, to be a terraced concavity in the solid globe of the sun as 

 under, with very steep sides, and breadths, c=l, b=2, c=3, the whole 



from a to a' being 13,500 miles, it will be obvious that the difference of the 

 levels of a and b cannot have so great a proportion to the sun's diameter 

 as the crater-walls of the moon bear to the diameter of that satellite, or 

 else cannot have anything like their steepness. The moon's crater-walls 

 are in height ^nn>- or even yirVo^ f the diameter, and often very steep. On 

 the sun the same proportion would give cliffs 400 or 800 miles high, and 

 with the spot in a position 60 from the centre of the sun, such cliffs would 

 on one side conceal half or more of one side of the terrace (b), while the 

 other side (b 1 ) of the terrace remained entirely visible. 



If we suppose that a is not a steep cliff but a prolonged slope, so that 

 even toward the edge of the sun the whole of the interior area may be 

 seen, the limit of the difference of level between the general surface (*) and 

 the interior terrace (b b') can be calculated. For example, 



In former observations of remarkable spots (1862), and again in 1863 



* Outlines of Astronomy (Ed. 1833), pi 3. d. 



