As early as 1981, Hans Freudenthal briefly mentioned Cauchy's work on "singular integrals (i.e., integrals of infinitely large functions over infinitely small paths [$\delta$ functions])" on page 135 of his biography of Cauchy in the Dictionary of Scientific Biography (edited by Gillispie).
A 1989 article by Detlef Laugwitz contains a section 5.5 starting on page 227 entitled "Cauchy and delta functions". At bottom of page 229, Laugwitz mentions Cauchy's use of the "language of infinitesimals" (on page 289 of Note XVIII in Cauchy's publication from 1827). This is in reference to Cauchy's infinitesimal $\alpha$ appearing in $\frac{\alpha}{\alpha^2+(\mu-a)^2}$, which integrated against $f(x)$ produces the value $f(a)$ of $f$ at the point $a$ (of course, from the modern point of view, the relation is not that of equality on the nose but rather infinite proximity).
Freudenthal's tone in 1981 seems to suggest that he is reporting something well-known. Which earlier historians (earlier than 1981) may have written about Cauchy's use of infinitesimals in writing down a formula for a delta-function (such as the $\frac{\alpha}{\alpha^2+(\mu-a)^2}$ above)?
