machinery,
which could suffer no injury from a blow of the head of a careless
observer. Strong and simple, they rarely got out of order. It is
said that an assistant who showed a visiting astronomer the transit
circle some times hit it a good slap to show how solid it was; but
this was not done on the present occasion. The little army had its
weekly marching orders and made daily reports of progress to its
commander, who was thus enabled to control the minutest detail of
every movement.

In the course of the evening Airy gave me a lesson in method, which
was equally instructive and entertaining. In order to determine the
longitude of Gibraltar, it was necessary that time signals should
be sent by telegraph from the Royal Observatory. Our conversation
naturally led us into a discussion of the general subject of such
operations. I told him of the difficulties we had experienced in
determining a telegraphic longitude,--that of the Harvard Observatory
from Washington, for example,--because it was only after a great deal
of talking and arranging on the evening of the observation that the
various telegraph stations between the two points could have their
connections successfully made at the same moment. At the appointed
hour the Washington operator would be talking with the others,
to know if they were ready, and so a general discussion about the
arrangements might go on for half an hour before the connections
were all reported good. If we had such trouble in a land line,
how should we get a connection from London to the Gibraltar cable
through lines in constant use?

"But," said Airy, "I never allow an operator who can speak with the
instruments to take part in determining a telegraphic longitude."

"Then how can you get the connections all made from one end of the
line to the other, at the same moment, if your operators cannot talk
to one another?"

"Nothing is simpler. I fix in advance a moment, say eight
o'clock Greenwich mean time, at which signals are to commence.
Every intermediate office through which the signals are to pass is
instructed to have its wires connected in both directions exactly
at the given hour, and to leave them so connected for ten minutes,
without asking any further instructions. At the end of the line
the instruments must be prepared at the appointed hour to receive
the signals. All I have to do here is to place my clock in the
circuit and send on the signals for ten minutes, comm