Vol. 70.] ANOTTEBSABY ADDRESS OF THE PRESIDENT. xciii 



observation. The transportation of material by rolling along the 

 bottom Avas dealt with in a different way. Various devices have 

 been suggested for determining its amount, and among them I 

 may mention a form of box-trap so constructed that it retained all 

 the gravel which the current might roll into it. By placing this 

 on the bottom of the river, it was believed by the inventor that an 

 estimate could be obtained of the amount of gravel travelling. 

 Such devices inspired me with profound mistrust — the initial error 

 appeared likely to be enormous, and it might be indefinitely 

 increased if the observations on a single spot were to be applied 

 to the length and breadth of the channel. 



We were fortunate, however, in obtaining, through the kindness 

 of the City Surveyor of Exeter, some highly significant records of 

 dredgings. The dredging has been conducted for several 3-ears in 

 a length of the river close above Exeter, which is held up at 

 a constant level by a weir, and is kept at a constant depth as a 

 bathing-place by removing the gravel as fast as it accumulates. 

 It may be assumed that no gravel is rolled over the weir. The 

 amount dredged, therefore, is the amount which the river has 

 rolled on to this part of its course. The records show that in the 

 seven years, 1904-10, 8071 cubic yards were dredged. 



Observations on the rainfall were provided by Dr. H. R. Mill, 

 Director of the British Rainfall Organization. The data upon 

 which calculations of the total annual fall are based consist of a 

 large number of daily observations on rain-gauges scattered quite 

 irregularly throughout the catchment-areas. Different methods of 

 dealing with this raw material will be presented by Dr. Mill and 

 Dr. Owens in our Report, and I will now merely point out one of 

 the most obvious difficulties that had to be met. For the keeping- 

 and reading of a gauge the volunteer services of a resident are 

 enlisted, but residents are apt to congregate in the most habitable 

 parts of a catchment-area and to avoid the moorlands. It thus 

 happens that the records from the valleys and lower reaches far 

 outnumber those from the uplands and upper reaches. Yet it is 

 upon the latter that the greater part of the rain falls. If, there- 

 fore, the average rainfall were calculated from the gauge-records 

 without taking into account the distribution of the gauges, a most 

 serious underestimate would ensue. 



For the determination of the limits of the river-basins it proved 

 necessary to examine the ground. The only map of river-basins 

 that was useful, is one published by the Ordnance Survey manv 



