Nearly a hundred years ago, Dr Heberden of London made the following experiment:—Having prepared three exactly similar rain-gauges, he placed one of them on the roof of Westminster Abbey, another on the roof of a neighbouring house, but at a much lower level, and the third in the garden of the same house. At the end of twelve months he found that the gauge on the roof of the Abbey had received 12·099 inches of rain, the gauge on the neighbouring house-top 18·139 inches, and the gauge on the ground 22·608 inches.
This paradoxical result required, of course, to be confirmed by other observers, and in other localities; and the similar results obtained by Dobson, Dalton, Howard, and especially by Arago at the Paris Observatory, and by Phillips at York, have amply supplied all that was wanting in this respect.