Because of a parameter that was assigned incorrectly in my
simulation, the numbers I just transmitted are not quite right.  They
apply to a lower rod length of 1 m, with the center of the x10 upper
mass a little below 5 cm above the axis.
     In the case of the 50 cm rod, the x 10 mass at just under 2.5 cm
yields a sensitivity at 20 s (Q = 100) of 39% that of the 99 m simple
pendulum.  Its sensitivity to high-frequencies is one-half of 1% of the
simple pendulum's.