128. Month-thirds and Market Weeks
Contrasting with this single diffusion of the seven-day week is the independent development in several parts of the world of other periods, marked either by sacred or secularly unlucky days or by markets or by divisions of the lunar month.
For instance, a ten-day week, having reference to the beginning, middle, and end of the lunation, was more or less reckoned with in ancient Egypt; ancient Greece; parts of modern central Africa; China, Japan, and Indo-China; and Polynesia. No historic connections are known between the custom in these regions; its official and religious associations are everywhere slender, and intervening nations either employ other periods or none at all. It looks, therefore, as if these might be cases of true parallelism, although in that event an American occurrence might also be expected and its absence seems in need of explanation. Moreover there is nothing very important about this reckoning; it is essentially a description of a natural event, and the only thing distinctive is its being threefold. If an institution as precise and artificial as our planetary week had been independently originated more than once, the fact would be more significant.
Regular market days among agricultural peoples have frequently led to a reckoning of time superficially resembling the week. Thus, in central Africa, south of the sphere of Islamic influences, markets are observed by a considerable number of tribes. Most frequently these come at four day intervals. Some tribes shorten the period to three days or lengthen it to five. Six, eight, and ten day periods appear to be merely doublings. The fairly compact distribution of this African market week points to a single origin.
The early Romans observed a regular eighth day market and semi-holiday. This might be connected with the African institution, but as yet cannot be historically linked with it.
In the less advanced states of Indo-China and many of the East Indian islands, even as far as New Guinea, five-day markets are the rule. This entire tract has many internal culture connections, so that within its limits diffusion has evidently again been active.
In ancient America, markets were customary every fifth day in Mexico, third day in Colombia, tenth day in Peru. These were also days of assembly and cessation from labor.
The American instances establish beyond cavil that some of these market weeks are truly independent evolutions. Moreover, they nearly all occur among peoples of about the same degree of advancement, at any rate on the economic side of their cultures. But it is only the idea, the outline of the institution, that is similar; its concrete cultural execution, as expressed in the length of the period, differs in Asia and Africa, and in the three American regions. That the Mexican and Southeast Asiatic weeks were both of five days, means nothing but the sort of coincidence to be expected when the choice of duration is limited to a small range, such as between three and ten days.