Alan, after thinking about it, what you say makes sense... but
somehow, possibly from vibration alone, the cap loosened over the course of a sailing season. I cant reliably explain how it did this (in hindsight, my original explaination may seem improbable as I stated it), but I can assure you it was tight at installation and worked its way loose all on its own, in less than one summer. Consider this: Mine is one of the thru hulls with the integrated self closing valve designed to minimize water flow when the sensor is removed. Also, the "notch" you speak of is actually a V shape that allows about 1/4" of radial play when the sensor is seated. Because its a V shape, as torque is applied to the sensor body, my assumption is that there would be an upward force on the sensor as its "key" rides up the side of the V notch. This force might well extent to the threaded cap. As the torque direction changes, the force on the sensor and/or cap would also change or at least be relaxed. I felt that with enough repetition of this process, the cap could loosen enough to eventually break a seal and cause a leak. Perhaps the mechanics of that explaination are incorrect. I dont need to debate it. I just know it can happen, because it happened to me.In any event, the Airmar folks (the manufacturer of the transducers Raymar supplies with their instruments, or at least, with my Bidata kit) must have forseen this problem as well. They include a wire safety wrap in the installation kit specifically for this purpose and explain its installation in the Airmar users manual. I stand by the reccommendation to use it.