WHAT FILLS THE STAR-DEPTHS? 
277 
ments might accordingly he looked for, since the exact reverse 
is the case. 
Now I conceive that so soon as we pass the third or fourth 
orders of star magnitude, we reach orders large enough, 
numerically, to supply the information, clear of the effects of 
mere accident, which we actually need in this instance. Among 
the stars down to the fifth magnitude, there is surely a suffi- 
cient number to enable us to begin to reason, with some degree 
of confidence, as to the constitution of stellar space. There- 
fore, when, in 1866, I was constructing my gnomonic maps of 
the heavens, in which stars of these orders are included, I was 
disposed to regard the signs I met with of special laws of dis- 
tribution as significant of real laws ; and accordingly I put 
forward, in that year, the theory that the stars are aggregated 
into streams and clustering aggregations, with relatively bare 
spaces all round them. And, furthermore, it seemed to me, 
even at that stage of the enquiry into the habitudes of stellar 
space, that the Milky Way probably consists of relatively 
minute stars, and not, as had been supposed, of stars generally 
comparable with our sun, and forming a system extending to 
enormous distances on all sides of us ; while I was led to regard 
the nebulae as belonging to the sidereal system, and not as ex- 
ternal galaxies resembling that system. 
But recently I have had occasion to apply processes of 
mapping to stars down to the sixth magnitude, or, in all, to four 
times as many stars as before. And clearly one cannot regard 
signs of arrangement among so many as 6,000 stars as being 
due to accident. The largeness of the number altogether pre- 
cludes the possibility of this being the case. 
When, therefore, it appears that among stars of the first six 
magnitudes there are signs of special laws of aggregation, we 
are bound to accept as legitimately following from the evidence, 
the conclusion that real laws of aggregation exist among the 
stars. We may not be able to tell what these laws are — we may 
mistake a number of separate clusters for a stream of stars, or 
the nearer end of a stream for the farther end, and so on ; but 
the broad fact remains that the stars are gathered into some 
regions and withdrawn from others, and, further, that within 
the same region of space stars of very different orders are, in 
many instances, gathered together. 
The general results of a systematic survey of the stars of the 
first six magnitudes seem certainly to force upon us such 
conclusions. They are as follows : — 
1. The southern hemisphere contains more stars of the 
orders considered than the northern, in the proportion of about 
seven to five. 
2. The stars of these orders are gathered into two definite 
