The eccentric is formed on the shaft coincident with the crank of the engine, so that the two arrive at their dead points simultaneously. The angular vibration of the line connecting the center of the eccentric with the trunnions of the link is the same as that of the connecting-rod. The connecting-rod of the length always used by me, namely, six cranks, makes the piston velocity at the head end of the cylinder 40 per cent. greater than at the crank end. By this construction the valve velocities were made to vary in the same ratio. A connecting-rod five cranks in length would increase this difference in piston velocities to 50 per cent., and one four cranks in length would increase it to 66 per cent. After Mr. Allen had explained his plan to me, I expressed my confidence that my governor would meet its requirements, and observed that it would enable a variable cut-off engine to be run as fast as a locomotive. Somewhat to my surprise, he replied that he wanted his cut-off compared with the liberating cut-off turn for turn; that it had an advantage which he thought would cause it to be generally preferred at the same speed. [Illustration: PORTER-ALLEN ENGINE. DIAGRAM OF ADMISSION-VALVE MOVEMENTS. RELEASE AND COMPRESSION ¹⁵⁄₁₆ OF THE STROKE ¹⁄₂ CUT OFF ¹⁄₂ CUT OFF ³⁄₈ CUT OFF ³⁄₈ CUT OFF ¹⁄₄ CUT OFF ¹⁄₄ CUT OFF ¹⁄₈ CUT OFF ¹⁄₈ CUT OFF TO VALVE AT CRANK END MEAN POSITION OF ROD RADIUS OF LINK TO VALVE AT HEAD END A. POINTS OF ADMISSION AND CUT-OFF. FOR DISTINCTNESS OF REPRESENTATION, THE THROW OF THE ECCENTRIC IS SHOWN ¹⁄₄ THAT OF THE CRANK. IN PRACTICE IT IS ONLY ¹⁄₁₂ THAT OF THE CRANK. The Diagram Drawn by Mr. Allen on his Engine-room Floor.] [Illustration: JOHN F. ALLEN]