The quest for measures that adequately show the aggregate development of prices and quantities through time, as well as what precisely should be understood by “adequate” started in the middle of the 19th century. Well known among the various indices are those proposed by Laspeyres in 1871, Paasche in 1874, and Fisher in 1922. Likewise well known is the principle, advocated by Lehr (1885) and Marshall (1887), of multiplying index numbers comparing adjacent periods to obtain measures covering longer time spans. At the beginning of the 20th century literally dozens of formulas were available, together with a number of criteria, also called tests, for choosing between them. All this was codified in Fisher's 1922 book The Making of Index Numbers.

Three years after the publication of this book the French economist Divisia (1925) presented a novel solution to the problem of splitting a value change into two parts, a part due to prices and a part due to quantities. He came up with two indices, an “indice monétaire,” which was a price index, and an “indice activité,” which was a quantity index. Both indices were defined as line integrals.

Up till this invention all price and quantity indices considered were essentially of the bilateral type; that is, they compared two time periods, employing price and quantity data pertaining to these two periods only. Chained indices are nothing but composites of bilateral indices. The novelty of Divisia's indices was that, as functions of continuous time, they take into account the prices and quantities of all, infinitely many, intermediate periods.