"Thank you," I said.

He drew the newspaper towards him and put it on his knee. He pulled out
a pencil; he held the foolscap up before his eye, and he began to
describe.

"The general principle upon which my speedometer reposes," he said
solemnly, "is the coordination of the cylinder and the cone upon an
angle which will have to be determined in practice, and will probably
vary for different types. But it will never fall below 15 nor rise over
43."

"I should have thought----" I began, but he told me I could not yet have
grasped it, and that he wished to be more explicit.

"On a king bolt," he said, occasionally consulting his notes, "runs a
pivot in bevel which is kept in place by a small hair-spring, which
spring fits loosely on the Conkling Shaft."

"Exactly," said I, "I see what is coming."

But he wouldn't let me off so easily.

"Yes, of course you are going to say that the whole will be keyed
together, and that the T-pattern nuts on a movable shank will be my
method of attachment to the fixed portion next to the cam? Eh? So it is,
but" (and here his eye brightened), "_anyone_ could have arranged
that. My particularity is that I have a freedom of movement even at the
lowest speeds, and an accuracy of notation even at the highest, which is
secured in a wholly novel manner ... and yet so simply. What do you