of a Stone laid on the Surface of the Ground. 



257 



mm. 



Feb. 19 0-00 



„ 24 0-28 



„ 29 0-43 



Mar. 7 0-54 



„ 14 0-97 



„ 22 1-43 



„ 28 1-69 



Apr. 6 0-89 



„ 18 1*11 



„ 25 1-43 



May 2 1 -89 



9 1-27 



May 18 3 



„ 23 3 



June 13 4 



„ 22 3 



„ 29 3 



July 12 3 



Aug. 22 4 



Sept. 7 . . . . 5 



„ 14 4 



„ 19 3 



„ 26 3 



Oct. 9 3 



The curve shown by the dotted line roughly represents the dampness 

 of the soil. Mr. Baldwin Latham has most kindly supplied me with 

 the rainfall during this period at Leaves Green, about 1 mile distant, 

 and nearly at the same level as Down. I have assumed that the soil 

 dries at a uniform rate ; this assumption cannot be correct, but no 

 other is possible. The varying rate of drying will, no doubt, depend 

 on temperature, wind, and dryness of the air, as well as on the rate at 

 which the water drains away. 



The ordinates are proportional to the amount of the rainfall, less the 

 assumed amount which has evaporated or drained away ; both quantities 

 are calculated from February 19, the date of the beginning of the 

 curve. The curves representing the dampness of the soil and the 

 movement of the stone are 16 mm. apart on February 19, the beginning 

 of the experiment, and the rate of drying has been assumed to be 

 great enough to bring them again 16 mm. apart on October 9, at the 

 end of the experiment. 



The curves follow each other in a striking manner after May 18. 

 On May 9 the stone-curve rises to a sharp peak when there was no 

 corresponding rainfall, suggesting an error in reading the micrometer 

 on that date ; this is the most probable explanation. Mr. W. N. Shaw 

 tells me that there was a thunderstorm on May 4 in the South and 

 West of England with variation in the local rainfall ; but this is unlikely 

 to be the explanation, as the rainfall between May 1 and May 9 at 

 Greenwich, 10 J miles distant, is the same as the Leaves Green, 1 mile 

 distant. On April 6 there is again a discrepancy ; the form of the 

 curve does not on this date suggest an error in the micrometer reading, 

 and no explanation is suggested. 



The direct effect of artifically wetting the ground was tried on 

 July 9, 1878. The ground was not dry, as there had been rain in 

 the previous night. About one hour after the water had been poured 

 on the ground near the stone it had risen 0*4 mm. ; six hours later it 

 had risen 0*1 mm. more. 



Fig. 3 shows the permanent downward movement of the stone from 



