572 Mr. Arthur Stanley Eddington [March 26, 



all the motion of the tail particles is directed nearly straight away 

 from the sun is sufficient to assure us that a solar repulsion, and not 

 the initial explosion, is predominant in causing the motion. The 

 suggestion that the sun might in some way act indirectly, causing 

 the explosion to expel the particles in the direction away from the 

 sun, cannot be accepted ; for the study of the envelopes, of which I 

 shall speak later, indicates that, if anything, the particles are projected 

 more abundantly towards the sun in the first instance. Besides, in 

 the case of the knot on October 3, if we prolong backwards the 

 uniform path we find it does not pass through the nucleus at all. 



There is a curious plate taken with the portrait-lens, which, 

 perhaps, represents something exceptional happening, but which 

 certainly confirms the idea that the repulsion does not continue to 

 act indefinitely. In it [shown on the screen] the tail matter is seen 

 to be no longer streaming off, but is hanging about in a cloud and 

 gradually diffusing. 



Perhaps the simplest way of explaining this unexpected absence of^ 

 acceleration is to suppose that there is some resisting material in the 

 space, so that a limiting velocity is reached for which friction and re- 

 pulsion counterbalance one another. Or we might, perhaps, suppose 

 that the bright patches do not strictly belong to the comet, but are due 

 to something projected from the sun, which causes a luminescence of 

 the tail matter of the comet through which it passes. I rather lean, 

 however, to an electrical explanation. It has been shown by various 

 writers that the repulsion which causes the tail may be due to the 

 action of the sun's electric field on the charged particles shot off from 

 the disintegrating head of the comet. It seems possible that these 

 might, after a short time, encounter particles of the opposite sign 

 and become neutralised. Assume for the sake of definiteness that 

 the sun has a positive charge, but is surrounded by the negative 

 electrons shot off from it, as from all white hot bodies, which form a 

 swarm extending beyond the comet. Then the positive particles of 

 the comet would be repelled, but at the same time they would be 

 bombarded by and encounter negative electrons which might combine 

 with and neutralise them, so that the repulsion would cease to act. 

 I think the recombination might take place much more easily where 

 the tail is densest, because a number of collisions would generally be 

 necessary to capture a negative electron, so that in the fine streamers, 

 which form the tails of ordinary comets, the repulsion might continue 

 to act much longer. This last is a suggestion, not so much based on 

 inherent probability, but put forward as a compromise to save us 

 from too hastily throwing over a great deal of research and study 

 on previous comets which has almost always assumed a constant 

 repulsion. 



Of the various theories that have been put forward in order to 

 account for the repulsion of comets' tails, besides the electrical 

 theories, probably the most popular is that which ascribes the 



