Interestingly, the distribution of Heine’s ‘archaeological’ bars does not match that of his radiocarbon dates: the centerpoints of two fall in the Colonial period, and one each in the Tlaxcala and Texcalac phases. On the basis of several dozen radiocarbon dates, I have documented the deposition of large volumes of alluvium between the Formative and Early Postclassic. In contrast, I have failed to positively identify any alluvium of Middle to Late Postclassic age. Downstream of Ladera, in an exposure of sandy near-channel

deposits with barely any sign of pedogenic development, a lens of charcoal buried at a depth of two meters yielded a date of 160 ± 40BP (Beta157074). This impinges on the end of the calibration data set, but the interval of highest probability at 1σ is AD1730-1780. The nature and stratigraphic context learn more of some other alluvia hint at CB-839 molecular weight a similarly recent date. At and east of La Laguna, many colluvial aprons grade into alluvial fan deposits. I have found several Postclassic and one apparently glazed sherd in them, but unfortunately in such low numbers and at shallow depths

that one cannot exclude intrusion from the modern ground surface. A cutbank of Los Ameyales is topped by more than a meter of bedded sands with no pedogenic imprint, likely derived from the erosion of the hillside of La Patada. Many barrancas (e.g., Concepción, Horcasitas, Coyotera) are bordered by ledges strewn with Middle to Late Postclassic sherds, sometimes on opposite banks of the same reach. Where the sherds are numerous, large and unabraded, excluding significant colluvial transport, this indicates that the stream has undergone an incision or major widening since the Postclassic. More precise dating of the onset new of incision, however, is often difficult, as exemplified by a cutbank at the foot of Loma La Coyotera. Its topmost palaeosol dates to 620 ± 50BP (Beta157070) and is buried by a wedge of colluvium. If the colluvium represents the activation of erosive processes in

the drainage that often precedes incision, the incision is more recent than the date. However, the span of the calibration (AD1280-1410 at 2σ), compounded by the uncertainty as to the residence time of the organic matter, and the delay in geomorphic response, mean that the stratigraphic sequence could be matched to any of rows A through E of Table 2. Sediment eroded off Las Margaritas now rests in a massive alluvial fan encroaching on and filling part of Lake Zacatepec. Postclassic sherds are present at the upper boundary of a soil buried on the lakeshore opposite the fan. This constitutes circumstantial evidence to link deposition to the abandonment of Las Margaritas, which I identify with Sacatepec, attested until at least the 1620s.