In the 1930s, Robert Horton analysed the data from his small experimental catchment at the Horton Hydrological Laboratory in Voorheesville, New York State. He also carried out a series of sprinkling infiltration experiments which suggest that infiltration rates over much of the catchment area set to meadow and grassland would be sufficient to infiltrate much of the rainfall, except under extreme rainfall conditions. The assumptions of his analysis are examined in detail and evidence for other explanations of the observed runoff coefficients considered. Finally, it is possible that the results of a plot experiment in 1932, for which runoff was measured continuously using a tipping bucket, are evidence that subsurface stormflow might be a significant runoff mechanism on this basin.