What my new therapist uses to prescribe in order to help healing not very severe tendon problems is slow eccentric exercises, aka negative phases, with very light weights, two or three sets, three times a day, everyday.
Now, I think that you don't need to be a rocket scientist to understand that, whatever is useful to heal and strengthen an injured tendon, is likely to be helpful in strengthening a healthy tendon as well.
Here are some links to places where I have seen sort of confirmation to this question, and references to scientific papers:
This link, tendinosis.org with this paragraph, among other interesting data:
Studies have shown that loading a tendon parallel to its length helps
the collagen fibers grow with better parallel alignment and speeds the
healing process. (...)
Some studies have shown that eccentric exercise is especially helpful
for tendinosis. [50, 51, 52, 53, 54] Eccentric exercise is when a
muscle is forced to lengthen while it contracts because it is being
used as a brake or to absorb energy while doing "negative work."
This other post in T-Nation, Accentuated Eccentric Training says:
Eccentric training is also a superior method to treat tendinitis when
compared with concentric exercise (Mafi et al. 2001). It could be
argued that this form of training is beneficial to injured athletes
and that it's relatively safer than concentric training even if the
loads used are greater.
But the best one on this question of tendons and tendonitis is this one (in T-nation too):
Surfing the web one day, I came across a posting on a newsgroup that
described a study done on runners with long-term, chronic Achilles
tendonitis. The study involved a new treatment that achieved
remarkable results in only 12 weeks. (...) The study, done in Sweden,
used two groups of fifteen middle-aged ex-runners who couldn't run due
to chronic Achilles tendonitis. (...) The special exercise program
involved performing calf raises with the legs both straight and
slightly bent twice a day, seven days a week for three sets of 15
repetitions (to hit both the gastros and soleus). The subjects
performed the concentric part of the exercise (raising the heel) with
the uninjured ankle and used the injured side to do the eccentric
lowering phase in a slow, controlled fashion. (...) At the end of the
twelve-week period, all the patients in the exercise group had
returned to running at their pre-injury level. (...) In a recent
two-year follow up of the fifteen runners in the exercise group,
fourteen were still running pain free(...)
Judge by yourself.