98 Proceedings of the British Association. 



of both hemispheres, I feel able to announce with confidence as a 

 general law, viz. that the character of easy resolvability into separate 

 and distinct stars, is almost entirely confined to nebulae deviating 

 but little from the spherical form ; while, on the other hand, very 

 elliptic nebulae, even large and bright ones, offer much greater 

 difficulty in this respect. The cause of this difference must, of 

 course, be conjectural, but, I believe, it is not possible for any 

 one to review seriatim the nebulous contents of the heavens without 

 being satisfied of its reality as a physical character. Possibly 

 the limits of the conditions of dynamical stability in the spherical 

 cluster may be compatible with less numerous and compara- 

 tively larger individual constituents than in an elliptic one. Be 

 that as it may, though there is no doubt a great number of 

 elliptic nebulae in which stars have not yet been noticed, yet 

 there are so many in which they have, and the gradation is 

 so insensible from the most perfectly spherical to the most elon- 

 gated elliptic form, that the force of the general induction is hardly 

 weakened by this peculiarity ; and for my own part I should have 

 little hesitation in admitting all nebulae of this class to be, in fact, 

 congeries of stars. And this seems to have been my father's opinion 

 of their constitution, with the exception of certain very peculiar look- 

 ing objects, respecting whose nature all opinion must for the present 

 be suspended. Now among all the wonders which the heavens 

 present to our contemplation, there is none more astonishing than 

 such close compacted families or communities of stars, forming 

 systems either insulated from all others, or in binary connexion, as 

 double clusters whose confines intermix, and consisting of individual 

 stars nearly equal in apparent magnitude, and crowded together in 

 such multitudes as to defy all attempts to count or even to estimate 

 their numbers. What are these mysterious families? Under what 

 dynamical conditions do they subsist ? Is it conceivable that they can 

 exist at all, and endure under the Newtonian law of gravitation with- 

 out perpetual collisions ? And, if so, what a problem of unimaginable 

 complexity is presented by such a system if we should attempt to 

 dive into its perturbations and its conditions of stability by the feeble 

 aid of our analysis. The existence of a luminous matter, not congre- 

 gated into massive bodies in the nature of stars, but disseminated 



