This article reexamines the thesis first put forward by Jean Bourgeois-Pich
at according to which the distribution of deaths in the first year of life
is governed by a universal law. An examination of data From a number of cou
ntries for both past and present populations, shows no such single and univ
ersal distribution, though it does reveal the existence of a number of dist
inct patterns which are subject to change over time.
The disparity between the observed data and those predicted by 'Bourgeois-P
ichat's law' cannot be taken as a systematic guide to the quality of the da
ta, although it may provide a clue to the problems affecting these data. La
stly, an examination of the distribution of infant deaths remains an effect
ive method for explaining the changes and variations occurring in the infan
t mortality rates of past and present populations.