and the Clouds of Magellan. 203 



bution of known nebulae may be best expressed thus : they are in 

 clouds, which do not lie in the plane of the star clouds, but on the 

 other hand, are not aggregated towards the poles of that plane. 

 In this view the greater cloud of Magellan is a cloud of peculiarly 

 rich constitution, including numbers of Milky Way objects, star 

 clusters, gaseous nebulae, and type V stars, as well as numbers 

 of nebulae which are probably spiral and non-Milky Way objects 

 (though of the spirality of any nebulae in this cloud there is no 

 real proof) ; but there need be no reason to suppose that it has 

 any special structural significance. 



This suggestion that the universe may consist of a number of 

 star clouds more or less in one plane, and a number of nebula 

 clouds out of the plane, is put forward merely as a working idea 

 about which facts may be grouped with some convenience. If it 

 has any advantage over the Spencerian view of the universe as 

 one organic whole, it is this, that we are not concerned to devise 

 an explanation of why stars and nebulae seem as a rule to avoid 

 one another, for we do not admit that the avoidance is so marked 

 as to make it probable that they are complementary members of 

 one system. 



It should however be noticed that the statistical methods 

 proper to the trial of the two hypotheses are totally different. If 

 the universe is to be discussed as one system, it is quite evident 

 that the fundamental plane is the plane of the Milky Way, and 

 that the general argument in statistical representations will be 

 galactic latitude ; galactic longitude will be of secondary im- 

 portance, and has, in fact, in many investigations been almost 

 disregarded. But as soon as we admit that the universe may 

 consist of a number of more or less independent clouds, an 

 attempt to find general laws with galactic latitude and longitude 

 as arguments becomes worse than useless, for it probably will 

 obscure the very discontinuities which are most significant. It 

 will be necessary to present the facts, not in the form of tables, 

 but of charts ; and the data which are charted must not be 

 smoothed, as has generally been done (notably in the distribution 

 charts of Schiaparelli, Seeliger, and Stratonoff), for the irregulari- 

 ties are probably all-important. 



The purpose of this note is therefore twofold : First, to suggest 

 a working hypothesis which seems to have some advantages over 

 that generally, though not perhaps always consciously, adopted in 

 discussions of star and nebular distribution; and secondly, to point 

 out that the statistics of the subject may be most conveniently 

 presented in a form which has been comparatively little used, that 

 of detailed distribution charts. 



