COMETS, THE SOLAR CORONA, AND THE AURORA. 49 



large, and throw out from them this appendage, which in- 

 creases in length and brightness till (as always happens in 

 such cases) they approach the sun and are lost in his beams. 

 After a time they again emerge on the other side, reced- 

 ing from the sun with a velocity at first rapid, but gra- 

 dually decreasing. It is, for the most part, after thus 

 passing the sun that they shine forth in all their splendour, 

 and their tails acquire their greatest length and develop- 

 ment, thus indicating plainly the sun's rays as the exciting 

 cause of that extraordinary emanation. As they continue 

 to recede from the sun their motion diminishes and their 

 tail dies away, or is absorbed into the head, which itself 

 grows continually feebler, and is at length altogether lost 

 sight of." 



Here, although unconsciously, Herschel has connected 

 the increase of brightness with the increase of speed with 

 which comets approach the sun, and the diminution in 

 brightness with the diminution of the velocity with which 

 they leave the sun. And although from HerscheFs remark 

 just quoted it might be inferred that proximity to the sun 

 is the cause of the increase of brightness, this is proved 

 not to be the case \ for (as in the case of Halley's comet) 

 when near its perihelion the tail sometimes dies away, and 

 the comet shrinks. In such cases, when the comet is 

 nearest to the sun there is no development of tail, which 

 shows clearly that it is not the intensity of the sun's rays, 

 but the change in their intensity, that is the exciting cause 

 of these extraordinary appearances ; so that there is no 

 reason to suppose that a planet composed of the same 

 material as a comet, no matter how close to the sun, would 

 show a vestige of tail or other cometic appearance. 



It is, then, to this change in position that we must 

 attribute those peculiar appearances which belong to 

 comets. 



Now is not electricity the very effect which would 

 SER. III. vol. v. E 



