On November 12, 1994, at 1430 eastern standard time, a Beech 35-C33, N5869S, owned and operated by Paul Millman, of New York, New York, lost power, and made an off airport forced landing in Coventry, Rhode Island. The airplane was destroyed and the pilots were not injured. Visual meteorological conditions prevailed and the flight was operated on an instrument flight rules (IFR) flight plan under 14 CFR Part 91. The instructional flight originated at Danbury, Connecticut, with the owner/operator occupying the left seat and the flight instructor in the right seat. The flight had been in cruise flight at 7000 feet and was descended to 3000 feet. An unsigned statement, attached to the NTSB Operator report stated: ...leveling off at 3000 feet, the engine failed...the prop speed was 2450 rpm (almost full forward), manifold pressure between 15 in. and 20 in. (we used 15 in. for descent and 20 in. for cruise), and fuel flow prior to failure was about 12 GPH... KF (flight instructor) flew the plane, trimmed for best glide, while PM (pilot under instruction) handled communication and attempted to restart the engine. The airplane came to rest in trees 1/4 mile short of runway 27. FAA Inspector Mr. Arthur Rica, an airworthiness inspector with the Boston Flight Standards Field Office, reported that when he examined the airplane after the accident, the electric boost pump switch was off and there was no emergency checklist available in the cockpit. When interviewed by telephone on November 15, 1994, the pilot was asked about the procedures used to restart the engine, the pilot said he richened the mixture, and switched tanks. He said he did not notice the fuel flow, nor did he mention that he turned on the boost pump. The unsigned, attached statement to the accident report also stated: Attempts to restart included switching fuel tanks, activating auxiliary fuel pump, mixture rich, throttle full forward, switched magnetos. According to the Pilot's operating handbook, the procedure for loss of engine power include checking the fuel flow and if it is low, place the mixture to rich and turn on the auxiliary fuel pump. If no change after a few moments, the auxiliary fuel pump is to be turned off. The engine was run while attached to the fuselage. Mr. Rica reported that the engine performed satisfactorily with the electric fuel boost pump, but would not run on the engine driven fuel pump. Examination of the engine driven fuel pump disclosed the drive shaft had failed at the shear neck. The bearings were seized and contained small pieces of aluminum, of undetermined origin. According to documents supplied by the FAA, the engine driven fuel pump was removed from N5869S and shipped to Approved Aircraft Accessories, Inc, Romulus, Michigan. It was received on April 26, 1994, with a note stating that no fittings were installed. It was overhauled and tested on April 27, 1994. According to the overhaul manual, fittings must be installed for testing of the pump. The pump was returned to the facility that shipped the pump, Bluebird Aviation Corp, Danbury, Connecticut, where it was reinstalled on N5869S on April 28, 1994. According to a signed statement from the mechanic who installed the pump, "...[The pump] was received bare without any fittings installed. The pump had accumulated 52.3 hours at the time of the accident.