On July 14, 1998, approximately 2210 mountain daylight time, a Polish-manufactured WSK PZL Mielec M18B agricultural aircraft, N4326M, operated by Hutchinson Spraying Service of Hingham, Montana, departed controlled flight and crashed during a landing attempt at the operator's private airstrip approximately 32 miles northwest of Havre, Montana. The airplane was substantially damaged and the airline transport pilot-in-command of the single-seat airplane was not injured. Visual meteorological conditions were reported at Havre at 2156 and no flight plan had been filed for the flight, which was operating under 14 CFR 91 at the time of the accident. The pilot reported to an FAA investigator that he had been spraying late since winds were unfavorable for spraying during the day, and that at the time of the accident, he was landing at his home airstrip after dispensing a load of parathion. He reported that in the landing attempt, he bounced twice and decided to go around. He stated that the aircraft then entered an uncommanded right roll, departed the airstrip to the right, touched its right wing tip to the ground, and crashed. The pilot reported to the FAA investigator that when the uncommanded right roll occurred, the stick was bound to the right. In his written report, the pilot noted that he noticed that the left-to-right movement of the control stick was stuck, and that he had applied great pressure on the stick to command a roll to the left. He said the control stick came unstuck and did move to the left, but the aircraft did not respond to "this extreme left movement of the stick. The aircraft continued in its turn/roll to the right. I pulled the throttle back to idle. The aircraft began to settle to the ground while in this turning motion. The aircraft went out of control and wrecked." The pilot added that "I felt that there was a complete loss of command of aileron or roll of the aircraft.... One thing is for sure. I was flying the airplane and I know how it reacted, and there are, in my opinion, mechanical malfunctions that support my description of how the aircraft flew. It was abnormal." Post-accident examination of the aircraft revealed both aileron bellcranks to be broken where they contacted the upper stops, with the right bellcrank locked in an aileron-up position. However, investigators found no evidence indicating fatigue failure when they inspected the bellcranks. The FAA investigator reported that the airplane had a total airframe time of 103 hours. Astronomical data indicated that at a point 32 nautical miles northwest of the Havre airport, sunset on the evening of the accident was at 2122, with end of evening twilight at 2257. The pilot's airstrip is unlighted.