29(5 Observations on the 



which they were observed. Hence, also, although by careful 

 observations they might probably have been discovered in 

 nearly all the phases, as in the present instance ; yet, in the 

 greater number of eclipses, they have been observed during the 

 continuance of greatest immersion only. From similar causes, 

 appearances resembling the above have been most conspicuous 

 in total eclipses, and might then, from a casual view, have 

 been easily mistaken for flashes of lightning. As an additional 

 proof, it may be worthy of remark that after attentively watch- 

 ing their progress, by taking into the field of the telescope that 

 part of the disc only where their appearance might be antici- 

 pated, these radiations were always found to come from the 

 periphery, never from the interior of the moon's orb. 



According to the quantity of the reflected rays, and their 

 being occasionally, from position, longer exposed to the view 

 of the observer ; those broader and more permanent streams of 

 light, which appear to have been visible in almost all eclipses 

 of this nature, would be produced. One of these, which oc- 

 curred nearly at the moment of conjunction, on the moon's 

 western limb, as illustrative of the situation of the lunar 

 mountains, has already been described. 



In reference to the present inquiry, it may be further re- 

 marked, that no haziness, or melting of the light into the sur- 

 rounding shade, could be observed ; on the contrary, the line 

 of demarcation was harshly traced, and the confines perfectly 

 distinct, which would hardly have been the case, had the illu- 

 minated surface been surrounded by a medium capable of re- 

 fracting the solar light. The reflected light, certainly, was less 

 vivid at a distance, from the circumference ; but this evidently 

 arose from its being divided into unequal streams by a ridge of 

 mountains. The outline of this ridge, as also that of the peri- 

 phery, were likewise very distinctly marked, without the slightest 

 appearance of an external and more faint illumination, such as 

 would have been produced by any attenuated luminous matter 

 surrounding them. (See Fig. 4, Plate ill, of last Number.) 



Another class of phenomena, from which considerable diver- 

 sity of opinion has originated, is that of detached masses, or 



