On Babbage’s 8 ystem of Mechanical Notation. 115 
possible for that particular purpose), what class of motion, 
whether crank, link motion, wheels, or cam work, he should 
adopt, and supposing he is required to adopt the more compli- 
cated form of cam work, the curve on the chart will enable him 
to sketch directly,and without any calculation the shape of his cum. 
(c.) Now as to the putting together of the machine. I do not 
mean by this, simply the reputting of the machine together after 
it has been once completed, but I mean the putting together for 
the first time of the various parts which have been separately 
and disjointedly completed to scale drawings, the ascertaining 
of the exact position in which to place each wheel, levers, and 
cams, so that all may work properly, and each part fulfil its 
proper duty, and at the proper time. In complicated machines, 
this is the one operation in which above all others, workmen are 
apt to make mistakes, for as the appliance for each motion has 
to be placed and “keyed” independently, it is very difficult indeed 
to say, except by a laborious tentative system of trial, whether 
that particular motion tallies perfectly with all the other motions, 
not only these already attached, but those to be attached. 
Here again Mr. Babbage comes to our assistance, for by 
attaching to the main shaft of the machine (which makes one 
revolution for each cycle), a cardboard disc divided into the same 
number of parts as the chart, we have in keying on any par- 
ticular motion, only to bring the shaft round till the divided disc 
reads such a division as corresponds to some certain action 
of that particular cam, or lever—-say the commencement or end 
of some particular action, and then turn the cam or lever on the 
shaft until that action does actually take place, and there key 
it, and sc on through all the motions: so that really the work- 
man might put the whole machine together, find out the position 
for all the motions, key them on, and be perfectly satisfied in 
his own mind, that the machine will work perfectly correctly, 
even though he never tried rt once. 
It may be said that a mathematician could describe and note 
down the nature and quality of all these actions without reference 
to a graphical representation, No doubt; but in the first place 
the mathematics required even for very simple machinery would 
~ be far higher than what we can expect mechanics to be conversant 
with, and J think we need no argument to show that a mathe- 
