SECT, xxxvi.] THE LIGHT OF COMETS. 409 



positions, to exhibit phases like the moon; but no such 

 appearance has been detected, except in one instance, when 

 they are said to have been observed by Hevelius and La 

 Hire, in the year 1682. In general, the light of comets is 

 dull that of the comet of 1811 was only equal to the tenth 

 part of the light of the full moon yet some have been bril- 

 liant enough to be visible in full daylight, especially the 

 comet of 1744, which was seen without a telescope at one 

 o'clock in the afternoon, while the sun was shining. Hence 

 it may be inferred that, although some comets may be alto- 

 gether diaphanous, others seem to possess a solid mass re- 

 sembling a planet. But whether they shine by their own 

 or by reflected light has never been satisfactorily made out 

 till now. Even if the light of a comet were polarized, it 

 would not afford a decisive test, since a body is capable of 

 reflecting light, though it shines by its own. M. Arago, 

 however, has, with great ingenuity, discovered a method of 

 ascertaining this point, independent both of phases and 

 polarization. 



Since the rays of light diverge from a luminous point, 

 they will be scattered over a greater space as the distance 

 increases, so that the intensity of the light on a screen 

 two feet from the object is four times less than at the dis- 

 tance of one foot ; three feet from the object, it is nine times 

 less, and so on, decreasing in intensity as the squares of the 

 distances increase. As a self-luminous surface consists of an 

 infinite number of luminous points, it is clear that, the greater 

 the extent of surface, the more intense will be the light; 

 whence it may be concluded that the illuminating power 

 of such a surface is proportional to its extent, and decreases 

 inversely as the squares of the distances. Notwithstanding 

 this, a self-luminous surface, plane or curved, viewed through 

 a hole in a plate of metal, is of the same brilliancy at all 

 possible distances as long as it subtends a ^ejisihlajfcngle, 

 because, as the distance increases, a greater portion comes 

 into view j and, as the augmentation/tisurface is as the square; 



