Maitchestcr Memoirs, Vol. I. (1906), No. f. 19 



particle would leave the sun altoi^ether. The scattering 

 of the particles is thus very rapid at distances greater 

 than two radii from the centre. But the same does not 

 hold within this distance. If we extend the series of 

 forces in the other direction by the numbers ii, 12, 13, 

 14, 15, the corresponding distances from the centre are 

 rSj, 171, 163, r56, I'SO : in other words, the inner 

 corona would be too nearly uniform, unless we have some 

 other source of variation. This we must seek in the 

 distribution of the sizes of the particles. We must suppose 

 that those which rise to considerable heights, because 

 light pressure nearly balances the sun's attraction, are 

 comparatively few, while there would be many for which 

 the balance was less complete. 



We have, therefore, to a certain extent, only replaced 

 one difficulty by another — or rather, only one enquiry by 

 another. Instead of looking for a cause for a certain 

 distribution of density, we now seek the reason of a 

 certain distribution of size. But I think that we have 

 advanced a step, although we may not have completely 

 solved the problem. That there should be this variation 

 in size near the point where light pressure nearly balances 

 attraction is of the nature of a vera causa. 



At the risk of appearing to argue in a circle, I now 

 call attention to the effect of this supposition on our 

 estimate of the absolute magnitudes of the velocities. If 

 we are right in regarding the variations of magnitude as 

 taking place in the force (and not in the velocity of 

 projection), then the average force is small, and the 

 velocities will also be small. This affords us a loophole 

 of escape from a possible difficulty. If particles were 

 -ejected with velocities of this magnitude, we might 

 reasonably expect the corona to change its form, at any 

 rate in detail, within a very short time. In one hour, for 



