Couple of things:
First, it was nice that in the recent movie that Holmes smoked a basket pipe as opposed to the iconic calabash. The calabash pipe was introduced by William Gillette in his portrayl of holmes, whereas it was never mentioned in Author Conan Doyles writings.
Second, based on the discussion above the research questions that appear to be asked are (1) could a pipe survive a fall from the height of the Richimbach Falls(sp) and if so could it be sufficently dried to be again usable (2) was temperature a factor. To accurately replicate the conditions of the fall Holme's pipe endured, we could drop pipes out of say a fifth story window into a basin of water, we would also need to replicate the turbulence of the river perhaps by continuly renewing the water with say a fire hose. We could monitor the initial temperature of the pipe with say a thermal couple and assume that it would equilibrate to the temperature of the water below. The absorption and desporption of water could be measured by weighing the pipe prior to filling, lighting and dropping; weighing the pipe when it is pulled out of the river, heat drying the pipe and again weighing (if all water came out the pipe should return to its original mass).
Third, the results of such an expereiment would have to be reproducible and statistically reliable (at least seven replicate measurements would have to be made per experimental run). Measurements would also have to be done under controlled conditions where only the varialbe studied could fluctuate. This means that we would need to use identical basket pipes. Potential pipe samples would need to be analyzed (gamma ray topography) to insure the absense of any hidden flaws that might produce a stress fracture and bias the results). For a given experimental run, each pipe would need to be dropped from the same height at the same intial temperature. For example to study the effects of temperature additional experimental runs (at seven pipes each) would have to conducted at other fixed temperatures. :puffpipe: