See other failures under 000.html.
Thomson Seatpost failed at transition from head to tube. From http://forums.mtbr.com/showthread.php?p=5579616#poststop, http://cozybeehive.blogspot.com/2009/04/thomson-elite-setback-post-failure.html and http://cozybeehive.blogspot.com/2009/04/broken-thomson-seatposts-overtightening.html as of 2010/03.
Rider was not injured, but rode approx. 25-30 km without a seat.
31.6 mm post; 410 mm length, 16 mm setback. User approx 90 kgf. 10 months and approx 3,000-3,500 km use, mostly on pavcement some on trails. Saddle was installed with the saddle back, clamped on the forward portion of the rails. Rider reports having installed using a torque wrench according to Thomson's instructions. Rider reports the seatpost was not damaged by e.g., driving in to a garage or clmaping the post in a stand.
The failed surface does not show "beach front" marks typical of long fatigue.
Note that a 31.6 mm post is larger than typical, 3,500 km use is unusually short for fatigue failures, and Thomson posts are widely used. These together suggest a manufacturing defect rather than a design defect.
Alternatively, a representative of Thomson examined the post and says there were divots in the post's cradle and the bolt heads were distorted. The divots reportedly only occur when there is substantial overtightening; and the distorted bolt heads take about 14 N-m (about 125 inch-pounds), which is about double Thomson's stated installation torque.
See also FAIL-132.html