On, October 9,1995, at 0930 hours mountain standard time, a Piper J3C-65, N59931, collided with trees following a power loss during the takeoff and initial climb from Ryan Field, Tucson, Arizona. The aircraft was owned and operated by the pilot and was beginning a local area personal flight at the time. No flight plan was filed and visual meteorological conditions prevailed. The aircraft incurred substantial damage. The certificated recreational pilot, the sole occupant, was not injured. The pilot stated that after completing a preflight check he entered runway 33 for a normal takeoff at full power. At an altitude between 50 and 100 feet, the engine lost power and the pilot began a gentle right-hand bank with intentions of landing in a grassy area parallel to the runway. The pilot further stated that he attempted to maintain flying speed and landed short of the intended grassy area. The aircraft sustained substantial damage to the left wing, landing gear, and fuselage. There was no postcrash fire. FAA airworthiness inspectors examined the aircraft and determined that the aircraft was utilizing automotive gasoline and there were 5 gallons of fuel available in the tanks. The fuel selector was found in the off position. The pilot stated that he never used the fuel shut-off and assumed that the airport people shut off the fuel on the day of the accident. The FAA inspector turned the fuel selector on and got good fuel flow at the carburetor. A postaccident engine run was conducted by FAA airworthiness inspectors. They found the that the engine started and operated at full rpm, then began to loose power output (300 to 400) rpm. The inspector applied the carburetor heat and the engine gained partial return of power. When the carburetor heat was removed from the engine, it returned to the to the previous running condition. The inspector stated that he suspected a float level problem and sent the carburetor out for analysis at an overhaul shop. The aircraft owner subsequently decided to replace the carburetor with a different type and the accident carburetor was lost in the overhaul process. Information Letter 8401 (12/15/84) from the Experimental Aircraft Association advises of a long standing problem with rubber tipped needle valves which swell when exposed to automotive fuels. The letter states that slow degeneration of inflight engine power has been observed when the rubber needle valve tips swell and decrease the orfice size at the carburetor inlet. Sudden losses of power have also been reported during takeoffs in aircraft following a long period of inactivity.