Tracing the first instance of a concept is not a certain or an always
fruitful exercise.  Hence, stating the first formal enunciation, not
necessarily mathematical, of a principle may be helpful in that quest.
What is the first statement of the principle of diminishing marginal
utility? The Australopithecus who had a fruit too many and said in these
pre-linguistic times "Man (not sapiens)! I am such a beast!" may be the
first precursor of the theory.

Mason Gaffney is correct that in ancient times philosophers condemned
greed and wealth for its own sake. Aristotle in his Politics argues that
trade is an unnatural mode of wealth-getting if exchange is not effected
in order to fill a certain need but in order to get more wealth, for there
is no end (telos) in such an endeavour.

Regarding diminishing marginal utility, its interest lies in using the
concept in order to derive a subjectivist theory of value.  JR Hicks
(1934) in an article on Walras' centenary (“Léon Walras”, Econometrica, 2
(4), pp. 338-348) writes about Auguste Walras, Leon's father:  he “was one
of those excellent people (they seem to have existed since very near the
dawn of history) who taught the true but unhelpful doctrine that value
depends on scarcity (rareté).” Diminishing marginal utility elevated that
concept on a higher analytical plane.

One can trace how you get a theory of value from utility in Jevons (1871)
discussion of Bentham's concept of utility. Jevons conflates all
dimensions of Benthamite utility to one to get his result. Gossen, Walras
and Menger, on the other hand, were not influenced by Bentham.

A more important question is whether the concept of "diminishing marginal"
comes from and to what uses has been put. One can start from Turgot's
Observation on a Memoire by Saint Peravy [1767] for the diminishing
returns on land and the notion of differential rent in English classical
political economy.  Then you have (1) notions of diminishing marginal
utility, (2) the creation of a notion of diminishing returns on capital
[See Pasinetti & Scazzieri in Palgrave 1987 "Capital Theory: Paradoxes"]
(3) an application of the notion to the product of labour in order to get
an ethical theory of distribution (JB Clark) and then (4) the suggestion
that somehow there is a relation between cost and quantity produced
[Sraffa, (1925/1997) “On the Relations between Cost and Quantity Produced”
in L.L Pasinetti (ed) Italian Economic Papers, Vol. 3, Oxford University
Press, Oxford, pp. 322-363)].

Regarding Mason Gaffney's comment on "mandatory pedantry", I think that at
these early stages contemporary formalistic orthopaedics did not apply.
Still, Bernoulli was a mathematician and Dupuit an engineer: mathematical
form was essential to them.  More interesting is the case of Gossen,
Jevons and Walras who were not well schooled in mathematics, but insisted
that their theory should assume a mathematical form if political economy
were to come of age.  Here we have self-imposed mandatory pedantry, a case
of "mind forg'd manacles".

Nicholas Theocarakis