ON THE SEICHES OF LOCH EARN. 395 



The pulley was supported on a horizontal board, perforated to receive it. Another 

 board, intended to hold the scale, was hinged to this at right angles ; it was hinged in 

 such a position that it could not open beyond the right angle. In this position it was 

 maintained by a rubber band. When not in use the two boards were folded together, 

 and enclosed the pulley and pointer between them. 



Index and Sccde. — Whatever size of pulley was used, the Index or Pointer was 

 made equal to four times the radius ; or the scale was so constructed as to indicate the 

 values at four times the radial distance, which comes to the same thing in practice. 

 There was thus an exaggeration of the seiches by four times, which made the smaller 

 ones easy to read. The scale was divided into inches and tenths of an inch. By using 

 paper also ruled in inches and tenths, the curves when plotted had exactly four times 

 the amplitude of the seiches. 



The Tripod. — Several camera-stands were tried. That found most convenient was a 

 light brass tubular telescopic tripod, the three legs fixed at the top to a circular plate, which 

 was joined to the table by a few turns of a single screw. It could be extended to a 

 length of 4 feet 6 inches, but was never used at more than 3 feet. 



The Well. — A very important part of the apparatus, and one which needed a good 

 deal of experimenting before it was brought to a satisfactory state. In calm weather 

 the tripod could be set up and the float placed directly in the loch, without any pro- 

 tection. This was, however, rarely possible, and we invariably used a zinc or tin 

 cylinder. A cylinder of 6-inch diameter was the smallest that would accommodate our 

 little floats and the syphon tube, leaving sufficient room for free motion. Ordinary 

 milk-cans supplied all that we desired. 



At first the water was admitted to the cylinder by a few small perforations in the 

 bottom, or low down at the sides. This did well enough, but had the disadvantage 

 that we did not know exactly what we were doing — we could not calculate the relation 

 of adit to water surface in the well. The method was soon abandoned for the syphon, 

 which was continuously used afterwards. The pail could be placed in the loch, but it 

 was there liable to disturbance if the wind rose ; and it was usual to put it in a larger well, 

 either completely cut off from the loch, or partly in connection with it and protected by 

 a breakwater. 



A large well, having free communication with the loch, was best. If the ingress 

 and egress of the water is not free enough, the breaking of the waves tends to keep 

 the well at a higher level than the loch and to float the cylinder. The same difficulty 

 occurs in another way when the well is quite isolated. The water percolating through 

 the sand raises the well above the level of the loch ; this was more marked if large 

 waves were breaking and saturating the soil around, though they might not enter the 

 well. The trouble from this cause was so frequent that we habitually weighted the 

 cylinders. If flat stones of the right size could be found, they were used ; if not, a layer 

 of sand or small gravel. The cylinder was also steadied outside by stones to prevent 

 the wind or the waves from moving it. 



