i 84 Haines. — A New A uxanometer. 
of fibre, the error introduced would be serious in accurate work. In order 
to compensate for the variability in the lengths of the fibres the arrange- 
ment employed is that seen in Fig. 2. In this figure it will be observed 
there is no longer a single continuous fibre from the plant apex to the pulley 
C as in Fig. 1, but there are three distinct fibres X 2 X 2 , Y, and z. The fibres 
x 2 and Y when they reach the pulley A pass round it a few turns and then 
become attached to its rim. Similarly the fibres Z and pass round the 
pulley B a few turns and are then also attached. In each case the two 
fibres run round the pulley in the same groove and in the same direction, 
but the pulleys A and B, instead of being simple, as in Fig. 1, are now 
double, each possessing two distinct grooves of the same radius and side by 
side (see inset, Fig. 2). In the 
second groove of each pulley there 
is attached a fibre running in the 
opposite direction to the other two, 
which when it leaves the pulley 
hangs down and carries a small 
weight. The weights V, v' carried 
on these fibres are exactly equal, 
and their function, which will be 
more fully understood later, is to 
maintain a tension in X x X 2 . The 
reason that v and v run in separate 
grooves on the pulleys A and B is 
to avoid friction with the other 
fibres, and to prevent the possi- 
bility of those which are being 
wound on to the pulleys becoming 
superposed on those which are 
leaving. The grooves are figured, 
for the sake of clearness (in Fig. 2), 
as if they were of different radii, 
but this is not actually so, their 
radii being, as already stated, exactly equal. As in Fig. 1, the pulley 
C is counterbalanced by the weight D, and, when the plant grows, is 
allowed to fall by the same distance as the stem apex rises. Now that the 
nature of the pulleys and connexions is understood it will be readily seen 
from the figure that stretching in the fibres Y and Z will tend to make the 
reading too large , while stretching in the fibre x t x 2 will tend to make the 
reading too small. Also stretching in Q R, which in this respect acts like 
a single fibre of length x (see Fig. 2), tends to make the reading too 
large. For compensation, therefore, the length of X x X 2 must be made equal 
to the sum of the lengths x } Y, and z. As the growth of the plant and fall 
Fig. 2. 
