g
one side of a parallel ruler of two or more points, g g, h h, i i.
Along i i is a slot and in this works a loaded block containing a
wheel P', whose plane is always parallel to i i. This block also
passes through a slot in D E, an arm at right angles to B C. A little
consideration will show that P', if worked at all, would trace out the
sum curve of P.

It was only when I showed the rough idea of this to Professor
Kennedy, with the view of ascertaining what would be the amount of
back-lash and friction, that I learned that Mr. Boys had already
invented a very similar integrator. In his model the double parallel
ruler is replaced by two endless strings and pulleys, and the bar, B
C, by a T square.

Although this integrator was afterward made in a less crude form, I do
not think it has ever been a practical instrument for the draughtsman.
Shortly afterward I came across a work by Abdank-Abakanowicz, entitled
"Les Integraphes," being a study of a "new kind of mechanical
integrator."

The new kind of integrator was really only an independent version of
Boys' instrument, but in many respects a great improvement. The real
merit will ultimately belong to the scientific instrument maker who
constructs an instrument reasonably cheap and capable of efficient
practical service. Abdank-Abakanowicz's integrator however certainly
went further in the practical direction than any previously
constructed. The drawing board machines, it is true, of rather a
complex nature, were actually exhibited to the Paris Academy, but no
more have been made. The instrument before me was made by Coradi, of
Zurich, on conditions laid down by me, namely, that the cost should
not exceed L14, and that polar distances should range between one and
ten half-inches. The first machine made by Coradi on these lines was,
by a misunderstanding, sold in Germany, but the one I exhibit is the
first, I believe, that has reached England, and to this extent I may,
perhaps, be permitted to call it new. I look upon it rather as a
suggestion upon which a still more practical instrument can be made in
this country than as a perfect model. I believe there would be a wide
sale for such an instrument were it once generally known to exist,
and, what is more to work efficiently. It remains for me to point out
in what the Abdank-Abakanowicz, or, rather, Coradi, integraph differs
from Boys' instrument.

Two points deserve special attention. In the first place, the fixed
f