322 The Construction of the Heavens. [July, 
used ‘‘to denote that some peculiar arrangement of stars in 
lines making different angles, directed to a certain aggrega- 
tion of a few central stars, suggested the idea that they 
‘(the former)’ might be in a state of progressive approach 
to them ‘(the latter).’ This tendency to clustering seems 
chiefly to be visible in places extremely rich in stars. 
In order, therefore, to investigate the existence of a clus- 
tering power, we may expect its effects to be most visible in 
and near the Milky Way.” I invite the reader’s special atten- 
tion to the circumstance that the Milky Way is here 
pointedly referred to as a stellar region distin@t in its 
characteristics from the region of stars forming our constel- 
lations. In studying Herschel’s papers we have to be 
continually on the watch for indications of the sort, since he 
does not always judge it necessary to make definite asser- 
tions of his opinions on such points. 
He then describes the irregular clusterings of stars, noting 
in particular, that ‘‘ though they are in general very pro- 
miscuously scattered, they are yet sufficiently drawn 
together to show that they form separate groups,” while 
in many places a falling off in the number of stars sur- 
rounding the clusters ‘‘indicates a tendency to future 
insulation.” ‘‘ Those which are in and very near the Milky 
Way,” he says, ‘‘may be looked upon as so many portions of 
the great mass drawn together by the action of a clustering 
power, of which they tend to prove the existence. In 
describing the various orders of irregular clusters, Herschel! 
is particular to notice vows and streams, ridges and shelvings, 
of stars, as indications of a preponderating clustering power. 
It is most important to notice here that Herschel had not 
yet replaced the theory of 1785 by any complete new 
theory.* He says at the close of the paper of 1814—“‘ The 
extended views I have taken in this and my former papers 
of the various parts that enter into the construction of the 
* The careful Struve did not fail to recognise the distinction between the 
theory of 1785 and Herschel’s subsequent labours. ‘Ou peut demander,” he 
says, ‘‘pourquoi les astronomes ont-ils maintenu généralement l’ancien 
systéme sur la Voie Lactée, énoncé en 1785, quoiqu’il ett été entierement 
abandonné par l’auteur luiméme comme nous l’avyons démontré. Je crois 
qu’il faut en chercher l’explication dans deux circonstances. C’était un 
systeme entier, imposant par la hardiesse et la précision géométrique de sa 
construction, et que l’auteur n’avait jamais révoqué dans sa totalité. Dans ses 
traités publiés depuis 1802, on ne rencontre que des vues partielles, mais qui 
suffisent, en les comparant entre elles, 4 comprendre l’idée finale du grand 
astronome.” I should point out, however, that Struve somewhat over-esti- 
mates the precision of the theory of 1785, and that he also fails to notice the 
circumstance that the papers of 1817 and 1818 are distiné in all respects from 
Herschel’s former work. 
