460 Sir George H. Darwin [April 26, 



many clusters. It is that of a number of lucid spots, of equal lustre, 

 scattered over a circular space, in such a manner as to appear 

 gradually more compressed towards the middle, and which com- 

 pression, in the clusters to which I allude, is generally carried so far, 

 as, by imperceptible degrees, to end in a luminous centre of an 

 irresolvable blaze of light. To solve this appearance it may be 

 conjectured that stars of any given very unequal magnitudes may 

 easily be so arranged, in scattered, much extended, irregular rows, 

 as to produce the above described picture ; or, that stars, scattered 

 about almost promiscuously within the frustum of a given cone, may 

 be assigned of such properly diversified magnitudes as also to form 

 the same picture. But who, that is acquainted with the doctrine of 

 chances, can seriously maintain such improbable conjectures ? " 



Later in the same paper he continues : — 



" Since then almost all the nebulfe and clusters of stars I have 

 seen, the number of -which is not less than three and tw^enty hundred, 

 are more condensed and brighter in the middle ; and since, from 

 every form, it is now equally apparent that the central accumulation 

 or brightness must be the result of central powers, we may venture 

 to affirm that this theory is no longer an unfounded hypothesis, l>ut 

 is fully established on grounds which cannot be overturned. 



" Let us endeavour to make some use of this important view of 

 the constructing cause, which can thus model sidereal systems. 

 Perhaps, by placing before us the very extensive and varied collec- 

 tion of clusters and nebula furnished by my catalogues, we may be 

 able to trace the progress of its operation in the great laboratory of 

 the universe. 



" If these clusters and nebulae were all of the same shape, and had 

 the same gradual condensation, we should make but little progress in 

 this enquiry ; but as we find so great a variety in their appearances, 

 we shall be much sooner at a loss how to account for such various 

 phenomena, than be in want of materials upon which to exercise our 

 inquisitive endeavours. 



" Let us, then, continue to turn our view to the power which is 

 moulding the different assortments of stars into spherical clustei's. 

 Any force, that acts uninterruptedly, must produce effects propor- 

 tional to tlie time of its action. Now^, as it has been shown that the 

 spherical figure of a cluster of stars is owing to central powers, it 

 follows that those clusters which, ceteris paribus^ are the most com- 

 plete in this figure, must have been tlie longest exposed to the action 

 of these causes. This will admit of various points of view. Suppose, 

 for instance, that 5000 stars had been once in a certain scattered 

 situation, and that other 5000 equal stars had been in the same situa- 

 tion, then that of the two clusters which had been longest exposed to 

 the action of the modelling power, we suppose would be most con- 



