V.
THE DROUGHT OF 1895-1896.
rainfall on the two catchment-grounds, is not necessarily identical each with the other, or with that at the Observatory.
I think however that the more probable explanation is that although the overflow ceased at Pokefulam on 28th August, it continued for some time in September at Taitam.
(5)
The facts observed during this drought are valuable, because the Taitam reservoir, at least, never overflowed, and therefore the available percentage of the rainfall can be calculated with precision, during the wet season. There is no information to show whether the same condition obtained at Pokefulam.
The rain-year of 1895-96 was probably the lowest on record, 45".8 as against 91".75 or about one half the general average. This defect is mainly observed in the wet season, May to August, during which the rainfall was 35".6 as against 60".34 or 59 percent only, of average. The rain in August was somewhat in excess of average but that of the other wet months considerably below it.
The rainfall of the dry season was also low, 18".23 as against an average of 31".39 or 58% of the average. Indeed it is the second lowest record for the dry season, of the whole period of 20 years.
It is unnecessary to repeat the calculations which I have made in the case of 1890-1. I gave these at length, in order to show whether I had fully comprehended the method of calculation, adopted by Mr Cooper.
The results as to the percentage of available rainfall may be thus tabulated.
6