He paused and indicated a number of little instruments on the table.

“For example,” he resumed, “the Lovibond tintometer tells me its story of the colour of the ink used in the ribbon of the machine that wrote this note as well as several standard specimens which I have been able to obtain from three machines on which it might have been written.

“That leads me to speak of the quality of the paper in this half-sheet that was found on Mr. Maitland.  Sometimes such a half-sheet may be mated with the other half from which it was torn as accurately as if the act were performed before your eyes.  There was no such good fortune in this case, but by measurements made by the vernier micrometer caliper I have found the precise thickness of several samples of paper as compared to that of the suicide note.  I need hardly add that in thickness and quality, as well as in the tint of the ribbon, the note points to person as the author.”

No one moved.

“And there are other proofs—­unescapable,” Kennedy hurried on.  “For instance, I have counted the number of threads to the inch in the ribbon, as shown by the letters of this note.  That also corresponds to the number in one of the three ribbons.”

Kennedy laid down a glass plate peculiarly ruled in little squares.

“This,” he explained, “is an alignment test plate, through which can be studied accurately the spacing and alignment of typewritten characters.  There are in this pica type ten to the inch horizontally and six to the inch vertically.  That is usual.  Perhaps you are not acquainted with the fact that typewritten characters are in line both ways, horizontally and vertically.  There are nine possible positions for each character which may be assumed with reference to one of these little standard squares of the test plate.  You cannot fail to appreciate what an immense impossibility there is that one machine should duplicate the variations out of the true which the microscope detects for several characters on another.

“Not only that, but the faces of many letters inevitably become broken, worn, battered, as well as out of alignment, or slightly shifted in their position on the type bar.  The type faces are not flat, but a little concave to conform to the roller.  There are thousands of possible divergences, scars, and deformities in each machine.

“Such being the case,” he concluded, “typewriting has an individuality like that of the Bertillon system, finger-prints, or the portrait parle.”

He paused, then added quickly:  “What machine was it in this case?  I have samples here from that of Dr. Boss, from a machine used by Mr. Masterson’s secretary, and from a machine which was accessible to both Mr. and Mrs. Maitland.”

Kennedy stopped, but he was not yet prepared to relieve the suspense of two of those whom his investigation would absolve.

“Just one other point,” he resumed mercilessly, “a point which a few years ago would have been inexplicable—­if not positively misleading and productive of actual mistake.  I refer to the dreams of Mrs. Maitland.”