105 
country and Dividing Range; in other words, they were not there to ringbark. But 
there has been much vegetation of a smaller kind, and it is believed by many that 
the eating-out and burning over of much of this vegetation is responsible to some 
extent for the changed condition of the western country to-day. 
The problem for New South Wales is to make the very best use of the water 
we receive, to keep it as long as we possibly can, and the encouragement of vegetable 
growth is a factor which tends to enable us to do this. 
The main forest covering of the mountains of Southern California consists of chaparral and brush. 
This covering holds the soil on the steep mountain sides, and detains the rainfall delivery, so that time is 
given for it to precolate into the water-veins and natural reservoirs. Where these water-sheds are burned 
over, the importance of the forest covering is at once demonstrated. In such districts the destructive 
force of the floods increase .... The rainfall is thus suddenly delivered to the injury of all. On 
the other hand, the perennial character of springs and streams is diminished or destroyed .... 
When the forest is gone on these steep sierras, floods and torrents alternate with wide and arid wastes of 
waterless torrent beds.* 
G. RAINFALL MEASUREMENTS IN FORESTS AND OPEN COUNTRY. 
We have large areas in this and neighbouring colonies where the forest is so thick that it will not 
pay to clear it away. Yet these very rain traps secure no more than the bare country, as I know by 
actual experiment carried on in one forest for a number of years, and in the dry time they suffer from 
drought just as the bare country does.f 
But the strongest argument adduced in the past to show the influence of forest on rainfall has 
existed is a comparison between rain-gauge measures in tlie forest and in the open field. Such records 
have been made for more than thirty years in France and Germany, and surely we must have here, if 
anywhere, a sufficient proof of a forest's influence. 
Admitting that we have perfect instruments and careful observers, there still remains a most serious 
doubt as to the immediate environment of each gauge, and as to the possibility of a direct comparison. It 
is probable that no two gauges 2,000 feet apart can be placed so as to catch the same amount of rain, 
though to all appearances the exposure is faultless in each case J 
Extreme caution is, therefore, needed in interpreting rainfall records in 
forests. We have also evidence of the partiality of rain showers on similar 
surfaces e.g., it sometimes rains in one paddock and not in an adjacent one. 
Professor Hazen gives instances of accurate records in forests and adjacent country 
by meteorologists, both in France and Germany, and shows the inconclusiveness of 
the results. Croumbie Brown, in his "Forests and Rainfall," also gives a full 
account of these researches, which cannot be further alluded to in detail here. 
Let me, however, point out that the humidity of a forest is not entirely a 
matter of rain-gauge measurements. I think, in order to thoroughly test this aspect 
of the question, the hygrometric state of the atmosphere in various places, whether 
carrying forests or other vegetation, and whether denuded by the hand of man or 
not, should be ascertained and compared for a series of years. The districts should 
be as numerous as possible, but we should not limit the observations to rain which 
can be measured in a rain-gauge. It will be found that much of the moisture which 
goes to assist pla'nt growth and to modify climate is not measured by such a crude 
instrument as that referred to. 
* Abbot Kinney, op. cit. See also article IX of my Fores-try scries, Agr'c. Ga:e!le, N.S.W., June, 19^0. 
t Mr. H. O. Russell, Hydney Morning Herald, 1st December, 1898. 
J Professor H. A. Hazen, op. cit. 
D 
