the whole unit.

[Illustration: FIG. 66]

[Illustration: FIG. 67]

In Fig. 66 A is the turbine spindle end and B the generator spindle end,
which it is required to drive. It will be seen from the cross-sectional
end view that both spindle ends are squared, the coupling C, with a
square hole running through it, fitting accurately over both spindle
ends as shown. Obviously the fit between the coupling and spindle in
this case must be close, otherwise considerable wear would take place;
and equally obvious is the fact than any want of alinement between the
two spindles A and B will be accompanied by a severe strain upon the
coupling, and incidentally by many other troubles of operation of which
this inability of the coupling to accommodate itself to a little want
of alinement is the inherent cause.

Looking at the coupling illustrated in Fig. 67, it will be seen that
something here is much better adapted to dealing with troubles of
alinement. The turbine and generator spindles A and B, respectively, are
coned at the ends, and upon these tapered portions are shrunk circular
heads C and D having teeth upon their outer circumferences. Made in
halves, and fitting over the heads, is a sleeve-piece, with teeth cut
into its inner bored face. The teeth of the heads and sleeve are
proportioned correctly to withstand, without strain, the greatest
pressure liable to be thrown upon them. There is practically no play
between the teeth, but there exists a small annular clearance between
the periphery of the heads and the inside bore of the sleeve, which
allows a slight lack of alinement to exist between the two spindles,
without any strain whatever being felt by the coupling sleeve E. The
nuts F and G prevent any lateral movement of the coupling heads C and D.
For all practical requirements this type of coupling is satisfactory, as
the clearances allowed between sliding sleeve and coupling heads can
always be made sufficient to accommodate a considerable want of
alinement, far beyond anything which is likely to occur in actual
practice. Perhaps the only feature against it is its lack of simplicity
of construction and corresponding costliness.

[Illustration: FIG. 68]

The type illustrated in Fig. 68 is a distinct advance upon either of the
two previous examples, because, theoretically at least, it is capable of
successfully accommodating almost any amount of spindle movement. The
turbine and generator spindle ends, A and B, have