"Yes," he replied, "there was, occasionally, very great danger
indeed. There is no accounting for the caprices of madmen; and, in my
opinion as well as in that of Dr. Tarr and Professor Fether, it is
never safe to permit them to run at large unattended. A lunatic may
be 'soothed,' as it is called, for a time, but, in the end, he is
very apt to become obstreperous. His cunning, too, is proverbial and
great. If he has a project in view, he conceals his design with a
marvellous wisdom; and the dexterity with which he counterfeits
sanity, presents, to the metaphysician, one of the most singular
problems in the study of mind. When a madman appears thoroughly sane,
indeed, it is high time to put him in a straitjacket."

"But the danger, my dear sir, of which you were speaking, in your own
experience -- during your control of this house -- have you had
practical reason to think liberty hazardous in the case of a
lunatic?"

"Here? -- in my own experience? -- why, I may say, yes. For example:
-- no very long while ago, a singular circumstance occurred in this
very house. The 'soothing system,' you know, was then in operation,
and the patients were at large. They behaved remarkably
well-especially so, any one of sense might have known that some
devilish scheme was brewing from that particular fact, that the
fellows behaved so remarkably well. And, sure enough, one fine
morning the keepers found themselves pinioned hand and foot, and
thrown into the cells, where they were attended, as if they were the
lunatics, by the lunatics themselves, who had usurped the offices of
the keepers."