Al'KIL I, 1922] 



|.. -,,„.,. 



^^H^ person is that so little is really known about 

 ^^H±i an everyday phenomenon as rainfall. It is a 

 satisfaction to remember that, thanks to the British 

 Rainfall Organization, more is known of the rainfall of 

 Britain than of any other country, but our complacency 

 ay be a httle disturbed when we reflect that for in- 

 vestigations as to what happens to the rain after it 

 falls we have to turn to other lands. Water-engineers, 

 indeed, have data from which much might be learned ; 

 but water engineers are secretive folk, and the records 

 of investigations on the run-off of the Severn, Exe, and 

 Medway remain the only records generally accessible. 

 The results of these investigations, though extremely 

 valuable, are not, however, very definite, as the areas 

 are so large and the problems correspondingly com- 

 plicated . More definite results are to be expected from 

 the experiment being carried out by the United States 

 Department of Agriculture in Colorado. ^ The areas 

 dealt with are small and the problem more defined, 

 though even in the small areas there studied conditions 

 are by no means so simple as might be desired. 



The intention of the experiment planned in 1909 was 

 to make a complete study of the effects of forest cover 

 on stream-flow and erosion. The main idea of the 

 method employed is simple enough. It was, to select 

 two small forest-covered valleys, contiguous, of the 

 same size, similar and similarly situated, to find the 

 rainfall and run-off from each, then cut down the forest' 

 from one of the areas and repeat observations. It 

 appears almost a laboratory experiment. The first 

 trouble was the trouble of the cook who desires to cook 

 a hare, or perhaps it would be better to say, a brace of 

 grouse, and it must be confessed at once that though 

 two somewhat similar birds were caught they were not 

 of the same kind, and as investigation proceeded un- 

 expected anatomical differences presented themselves, 

 extremely interesting in their own way, but not making 

 for uniform cooking ; it appears also that even if they 

 had been both of a kind they were particularly difficult 

 birds to cook. The valleys chosen lie about the 10,000 

 feet level in a region with precipitation about 20 inches 

 a year, about half of which falls as snow and a goodly 

 proportion of the rain in thunderstorms, both pheno- 

 mena introducing difficulties. 



The publication before us is a preliminary report 

 giving an account of the first part of the experiment 

 from 191 1 to 1919, and discusses the data obtained 

 while both valleys, A and B, were forest covered. 

 Both valleys are small, B of 200 acres and A a little 

 larger, varying in elevation from just over 9000 feet to 

 just under 11,000 in the case of B, and somewhat over 

 in the case of A. The geological structure is identical, 

 namely, augite-quartz-latite, little porous to water, 

 covered with a few feet of soil and decomposed rock, 

 porous and sandy in texture, forming a permeable and 

 well-drained top layer. The forest cover, conifers of 

 various kinds, is almost identical. The valleys are not, 

 however, of quite the same shape — A is long and narrow, 

 B is much more like a bowl : the exposure is rather 



^ * " Stream-flow Experiment at Wagon Wheel Gap, Colorado," Monthly 



Weather Bureau Supplement, No. 17. Government Printing Office, Washing- 

 ton, 1922. 



NATURE 



417 



Forests in Relation to Stream-flow and Erosion. 



different, the centre line of A being south of east, while 

 that of B is north of east. This is important in view 

 of the fact that the winter snowfall runs off as it is 

 melted by the summer sun, and indeed both the time 

 and degree of response of the two streams to any factor 

 influencing the regime are somewhat dissimilar. For 

 example, after rainfall A rises more rapidly and reaches 

 its maximum flow earlier than B, B may then be higher 

 than A for a time, while at the end of the flood A may 

 be higher than B. As a result, it has been necessary to 

 construct tables and diagrams to show the relation of B/A 

 for a great variety of conditions, and some i6 " rules " 

 have been formulated for comparing the discharge of 

 B when the discharge of A and the rainfall is known. 



The readings for the run - off may probably be 

 accepted. Very great care has been taken to construct 

 suitable dams, gauges, and basins. The construction of 

 the measuring apparatus is described in great detail, 

 and the readings appear in general to have been exceed- 

 ingly accurate and trustworthy. But it is a little difficult 

 to place implicit confidence in either the precipitation 

 statistics or the use that is made of them. Though 

 details are, perhaps significantly, lacking, it is evident 

 that the exposure of the gauges for rain and snow is 

 not up to the standard required in this country, while 

 their distribution also leaves something to be desired. 

 Only five were set up in the two valleys ; two are close 

 together in the lower part of each basin and one at 

 almost the highest point of A, while a sixth was just 

 outside the lower portion of both basins. 



The number would, of course, be abundant for ordinary 

 rainfall work, but in a scientific experiment which is 

 otherwise marked by accuracy, British experience would 

 suggest that the number was inadequate, and we should 

 imagine that over a vertical height of 2000 feet there 

 would be considerable differences in rainfall, especially 

 when a good proportion of the rain falls in thunder- 

 storms. It is possible that conditions are different in 

 Colorado, but we should have been more satisfied if 

 evidence had been adduced to show that this was so. 

 Nor is our confidence increased when we learn that in 

 the second part of the experiment, when the forest is 

 removed from B, only the gauges in the A valley are 

 to be read. It is scarcely sufficient to say that " the 

 use of the single record cannot be seriously objected 

 to when it is considered that at the lower end of A 

 there is the choice of the better catch of two gauges, and 

 this value is averaged with the catch of the third gauge 

 at the head of the valley." The italics are ours. It is 

 only fair to say that much more care has been taken 

 with another and equally important side of the problem, 

 the melting of the snow. Observations of the depth of 

 the snow at the time of thaw are taken at a considerable 

 number of points. 



It will be interesting to see in ten years' time the 

 results of removing the forest. No doubt valuable 

 results will be obtained which will be of use in dealing 

 with the Forest Reservations of the Rockies, but even 

 so, light will be thrown on only a small portion of the 

 small problem. We shall know what is the effect of 

 removing forest cover only under somewhat special 

 conditions. There will be plenty room for further 

 investigation. 



NO. 2735, VOL. 109] 



