On June 3, 2000, about 1510 Pacific daylight time, a Bell 212, C-FHDY, registered to and operated by Alpine Helicopters as a 14 CFR Part 91 local pilot proficiency evaluation flight, experienced an in-flight tail rotor system failure while practicing autorotations to a hover at the Eugene Airport, Eugene, Oregon. Visual meteorological conditions prevailed at the time and no flight plan had been filed. The helicopter was substantially damaged. The two airline transport pilots were not injured. The flight originated from Eugene about one hour prior to the accident. The pilot-in-command (PIC) reported that he is a check airman for the company and was giving the other pilot a pilot proficiency check. The PIC stated that during the third practice power recovery autorotation to a hover, at about 600 feet above ground level, they heard a loud bang, followed by a severe vibration. The PIC lowered the collective all the way and rolled off both throttles, while the other pilot transmitted a mayday. The PIC continued the autorotation and the helicopter touched down without further incident. The PIC reported that after the accident, they found that the 90-degree gearbox had separated from the tail boom. The gearbox, with tail rotors attached, was later found approximately 250 yards from the helicopter touch down point. Maintenance records, obtained from the owner, indicate that the tail rotor blades, yoke assembly, hub and blades, and hub assembly was overhauled by the owner on November 3, 1999. The helicopter had accumulated a total time since overhaul of 701.9 hours. Inspection of both tail rotor blades revealed that the tip weight block from one of the blades was missing along with the end closure. About four inches of skin, starting from the leading edge was peeled outwards on both sides of the blade. The blade was designated as Blade A, s/n: A-9495, and was intact except for the tip damage. The other blade was designated as Blade B, s/n: A-9628, and was bent in half about 90-degrees and exhibited evidence of ground impact. The tip weight block remained intact in this blade. A representative from Bell Helicopters reported that Blade A was manufactured in November 1994, and Blade B in April 1995. Both blades were purchased from Bell Helicopters in October 1995. The maintenance records indicate that the blades had been installed on two different helicopters before being installed on C-FHDY in May 1998. Both blades (s/n: A-9495 and s/n: A-9628) had accumulated a total time since new of 2,918.9 hours. The records indicate a tail rotor blade retirement life of 5,000 hours. The tail rotor blades were sent to Transportation Safety Board of Canada, Material lab for inspection. The specialist reported that remnants of adhesive remained close to the trailing edge, and attached to the honeycomb core surface. Fewer traces of the adhesive remained on the skin surfaces and in the region of the tip block. The representative from Bell Helicopters reported that the adhesive bonding material is the primary securing method for the tip block. The secondary securing method consists of four countersunk screws in line chordwise and offset in pairs on each side of the blade. A visual inspection of all four holes showed no significant evidence of deformation. A field repair of the original factory painted surface on the upper surface, aft hole was also noted. An absence of adhesive in the bonding area for the end closure was also noted. The quality of the bond between the blade skin and the spar cap was examined. The examination revealed that 75-85% of the surface separated as a result of an adhesive debond between the adhesive and the spar surface. Evidence of staining was noted around the screw holes, with soot intrusion. The examiners reported that, "The separation was likely a result of environmental affects of moisture on the bond between the adhesive and the spar surface." The condition of the countersunk holes was also examined. After measurements were made, it was found that the diameters had been drilled larger than the manufacturers drawing requirements. (See attached Engineering Reports) The countersunk angle was calculated and also found not to meet the drawing requirements. The examiners reported that as a result of the countersunk holes being manufactured too large, it allowed the four secondary load path screws to slip behind the spar, resulting in the tip block separating from the blade. The representative from Bell Helicopters reported that although the bonding adhesive is identified as the primary securing method for the tip block, and the four countersunk screws as the secondary, either method independent of the other is capable of carrying the load.