22

egeson’s weather system.

positions of the areas drawn upon a small map, and preservedfor reference.

For a few years the daily reports were thus treated, thatperiod being thought sufficient for the determination of therelation of clouds to rain, and of rain as observed at nine o’clocka.m. to the rainfall during twenty-four hours, as well as theeffect of the winds upon these. For Sundays and holidaysthe rainfall was always included in the following day, and bya process of interpolation their values were determined fromthese. For rain and clouds a mean of Saturday and Mondaywas taken to represent Sunday.

For the remainder of ten years the areas were determinedfor rainfall only. This brings us to 1878. The number ofstations had by that time decreased to forty, and were dis-tributed very unfavorably for the further continuance of theestimation of area in that way. Twenty representativestations were then selected as nearly equidistant as possible.The number of stations reporting rain each day was dividedby two to give the area in tenths of the whole. This wasdone until 1870. Prior to that year the stations were few innumber, and in 1858 there were only six available. Still, theprocess was applied to those years, but the result, of course,only refers to so much area as the stations could be fairlytaken to represent. In 1858 the area was thus only abouttwo-tenths of the whole colony. It is necessary that thusmuch should be explained in detail to enable the reader tojudge of the value of the little diagram (representing thevariability in the time of arrival of the four rain periods forthe whole colony, as deduced from the areas of rainfall)given at the foot of a page of diagrams already described,but to which we shall have occasion to refer again, as wellas to show how far dependence may be placed upon thedeductions which are to follow.

The areas determined, the time-scale is again brought intorequisition, as in the case of the individual stations and of thesun. Of the thirty-one years, I again select two, 1887 and1888. They are representative of extreme conditions, inas-