rovocative of internal dissension.
But this cause has not operated in the case of Holden's successors.
Keeler became the second director in 1897, and administered his office
with, so far as I know, universal satisfaction till his lamented
death in 1900. It would not be a gross overstatement to say that his
successor was named by the practically unanimous voice of a number
of the leading astronomers of the world who were consulted on the
subject, and who cannot but be pleased to see how completely their
advice has been justified by the result of Campbell's administration.

VIII

THE AUTHOR'S SCIENTIFIC WORK

Perhaps an apology is due to the reader for my venturing to devote a
chapter to my own efforts in the scientific line. If so, I scarcely
know what apology to make, unless it is that one naturally feels
interested in matters relating to his own work, and hopes to share
that interest with his readers, and that it is easier for one to write
such an account for himself than for any one else to do it for him.

Having determined to devote my life to the prosecution of exact
astronomy, the first important problem which I took up, while at
Cambridge, was that of the zone of minor planets, frequently called
asteroids, revolving between the orbits of Mars and Jupiter. It was
formerly supposed that these small bodies might be fragments of a
large planet which had been shattered by a collision or explosion.
If such were the case, the orbits would, for a time at least, all
pass through the point at which the explosion occurred. When only
three or four were known, it was supposed that they did pass nearly
through the same point. When this was found not to be the case,
the theory of an explosion was in no way weakened, because, owing
to the gradual changes in the form and position of the orbits,
produced by the attraction of the larger planets, these orbits would
all move away from the point of intersection, and, in the course of
thousands of years, be so mixed up that no connection could be seen
between them. This result was that nothing could be said upon the
subject except that, if the catastrophe ever did occur, it must have
been many thousand years ago. The fact did not in any way militate
against the theory because, in view of the age of the universe, the
explosion might as well have occurred hundreds of thousands or even
millions of years ago as yesterday. To settle the question, general
formulae must be