1921.] Birks and Bates.—Annual Rainfall Fluctuations. 
175 
engineer to make his own forecast of the future more intelligently than 
without them, and the figures therefore have this amount of value. 
These figures being based only on annual rainfall fluctuations, care 
must be taken not to use them in any way in connection with flood- 
occurrences or temporary droughts lasting only for a short period of the 
year. The frequency of the occurrence of such events in the past can 
be determined only by a much fuller investigation of the figures for much 
shorter intervals than the yearly periods. 
Kg. 2 gives the records available for five other stations, eight in all, 
for the past thirty years, to which similar principles might be applied, but 
owing to the short period for which records are available in these additional 
cases the results would not have the same value. 
As far as these figures go, the amount of the annual rainfall is apparently 
quite erratic, and follows no cyclic variation. But in other places the 
eleven-year sun-spot cycle is distinguishable, and even a three-year cycle. 
Before this can be definitely determined under local circumstances a much 
fuller investigation will be required, which it is hoped will be undertaken 
in due course. For this purpose shorter periods should be taken, and 
accidental fluctuation eliminated -by taking special, series of averages. We 
are indebted to Professor Marsden for the proposal that this might be a 
suitable investigation for a post-graduate science student, and, amongst 
other suggestions, that the use of consecutive yearly means from month 
to corresponding month a year ahead and by taking the smoothed mean 
———-—offers a likely method of minimizing the effect of personal 
or accidental errors and obtaining a truer rainfall curve. Moreover, the 
question of the sites of the rain-gauges in each locality requires careful 
checking if these have been moved during the period of observation. In 
each place the gauges have been moved, and records are required to 
correlate the rainfalls of the old and new sites. A fuller investigation 
is being made of this matter by the Dominion Meteorologist, and it is 
hoped that it will be possible to publish details shortly. In the meanwhile 
it should be noted that it is proposed to again move the Wellington 
meteorological station to a new site, and it is hoped that a new site will 
