% This paper has been transcribed in Plain TeX by
% David R. Wilkins
% School of Mathematics, Trinity College, Dublin 2, Ireland
% (dwilkins@maths.tcd.ie)
%
% Trinity College, 1st June 1999.
\magnification=\magstep1
\vsize=227 true mm \hsize=170 true mm
\voffset=-0.4 true mm \hoffset=-5.4 true mm
\def\folio{\ifnum\pageno>0 \number\pageno \else\fi}
\font\Largebf=cmbx10 scaled \magstep2
\font\largerm=cmr12
\font\largeit=cmti12
\font\sc=cmcsc10
\pageno=0
\null\vskip72pt
\centerline{\Largebf ON DIFFERENCES AND DIFFERENTIALS OF}
\vskip12pt
\centerline{\Largebf FUNCTIONS OF ZERO}
\vskip24pt
\centerline{\Largebf By}
\vskip24pt
\centerline{\Largebf William Rowan Hamilton}
\vskip24pt
\centerline{\largerm (Transactions of the Royal Irish Academy,
17 (1837), pp.\ 235--236.)}
\vskip36pt
\vfill
\centerline{\largerm Edited by David R. Wilkins}
\vskip 12pt
\centerline{\largerm 1999}
\vskip36pt\eject
\null\vskip36pt
{\largeit\noindent
On Differences and Differentials of Functions of Zero.
\hskip 0pt plus1em minus0pt
By {\largerm WILLIAM R. HAMILTON}, Royal Astronomer of Ireland.}
\nobreak\bigskip
\centerline{\vbox{\hrule width 72pt}}
\nobreak\bigskip
\centerline{Read June~13, 1831.}
\bigbreak
\centerline{[{\it Transactions of the Royal Irish Academy},
vol.~17 (1837), pp.\ 235--236.]}
\bigbreak
The first important researches on the differences of powers of
zero, appear to be those which Dr.~{\sc Brinkley} published in
the Philosophical Transactions for the year 1807. The subject
was resumed by Mr.~{\sc Herschel} in the Philosophical
Transactions for 1816; and in a collection of Examples on the
Calculus of Finite Differences, published a few years afterwards
at Cambridge. In the latter work, a remarkable theorem is given,
for the development of any function of a neperian exponential, by
means of differences of powers of zero. In meditating upon this
theorem of Mr.~{\sc Herschel}, I have been led to one more
general, which is now submitted to the Academy. It contains
three arbitrary functions, by making one of which a power and
another a neperian exponential, the theorem of Mr.~{\sc Herschel}
may be obtained.
Mr.~{\sc Herschel's} Theorem is the following:
$$f(e^t) = f(1) + t f(1 + \Delta) o^1
+ {t^2 \over 1 \mathbin{.} 2} f(1 + \Delta) o^2
+ \hbox{\&c.}
\eqno {\rm (A)}$$
$f(1 + \Delta)$ denoting any function which admits of being
developed according to positive integer powers of $\Delta$, and
every product of the form $\Delta^m o^n$ being interpreted, as in
Dr.~{\sc Brinkley's} notation, as a difference of a power of
zero.
The theorem which I offer as a more general one may be thus
written:
$$\phi(1 + \Delta) f \psi(o)
= f(1 + \Delta') \phi(1 + \Delta) ( \psi(o) )^{o'};
\eqno {\rm (B)}$$
or thus
$$F(D) f \psi(o) = f(1 + \Delta') F(D) (\psi(o) )^{o'}.
\eqno {\rm (C)}$$
In these equations, $f$, $\phi$, $F$, $\psi$, are arbitrary
functions, such however that $f(1 + \Delta')$, $\phi(1 +
\Delta)$, $F(D)$, can be developed according to positive integer
powers of $\Delta'$ $\Delta$ $D$; and after this development
$\Delta'$ $\Delta$ are considered as marks of differencing,
referred to the variables $o'$ $o$, which vanish after the
operations, and $D$ as a mark of derivation by differentials,
referred to the variable~$o$. And if in the form (C) we
particularise the functions $F$, $\psi$, by making $F$ a power,
and $\psi$ a neperian exponential, we deduce the following
corollary:
$$D^x f(e^o) = f(1 + \Delta') D^x e^{o'}
= f(1 + \Delta') o'^x;$$
that is, the coefficient of
$\displaystyle {t^x \over 1 \mathbin{.} 2 \, \ldots \, x}$
in the development of $f(e^t)$ may be represented by
$f(1 + \Delta) o^x$; which is the theorem (A) of
Mr.~{\sc Herschel}.
\nobreak\bigskip
June~13, 1831.
\bigbreak
\centerline{ADDITION.}
\nobreak\bigskip
The two forms (B) (C) may be included in the following:
$$\nabla' f \psi(o') = f(1 + \Delta) \nabla' ( \psi(o') )^o.
\eqno {\rm (D)}$$
To explain and prove this equation, I observe that in
{\sc Maclaurin's} series,
$$f(x)
= f(o) + {D f(o) \over 1} x
+ {D^2 f(o) \over 1 \mathbin{.} 2} x^2
+ \cdots
+ {D^n f(o) \over 1 \mathbin{.} 2 \, \ldots \, n} x^n
+ \cdots $$
we may put $x = (1 + \Delta) x^o$ and therefore may put the
series itself under the form
$$f(x)
= f(o) + {D f(o) \over 1} (1 + \Delta) x^o
+ {D^2 f(o) \over 1 \mathbin{.} 2} (1 + \Delta)^2 x^o
+ \hbox{\&c.}$$
or more concisely thus
$$f(x) = f(1 + \Delta) x^o:
\eqno {\rm (E)}$$
which latter expression is true even when {\sc Maclaurin's}
series fails, and which gives, by considering $x$ as a
function~$\psi$ of a new variable $o'$ and performing any
operation $\nabla'$ with reference to the latter variable,
$$\nabla' f \psi(o') = \nabla' f(1 + \Delta) ( \psi(o') )^o.
\eqno {\rm (F)}$$
If now the operation $\nabla'$ consist in any combination of
differencings and differentiatings, as in the equations (B) and
(C), and generally if we may transpose the symbols of operation
$\nabla'$ and $f(1 + \Delta)$, which happens for an infinite
variety of forms of $\nabla'$, we obtain the theorem (D). It is
evident that this theorem may be extended to functions of several
variables.
\nobreak\bigskip
June~20, 1831.
\bye