i894. THE EVOLUTION OF THE THAMES. 105 



Some, moreover, occur in districts where it is probable that this 

 formation never extended. 



The distribution of the Sarsens enables us to approximately 

 determine the original extension of the Tertiary Sands that yielded 

 them. The Sarsens are abundant on both sides of the Kennet valley, 

 on the Marlborough Downs, and thence along a line past Avebury 

 north to Swindon. The blocks here are small, as they are also at 

 Highworth and Hannington. They occur on the dip slope of the 

 Chalk of the Berkshire Downs as far as Aldbourne and Lambourne, 

 but they appear to be entirely absent from the higher part of the 

 Dov/ns. In a walk from Ashbury to Aldworth, along the edge of the 

 Chalk escarpment past Uffington Castle and White Horse Hill, not 

 a single Sarsen stone was seen, except those forming the cromlech 

 and circle of Wayland Smith's cave and the famous " Blowing Stone." 

 These are certainly not in situ : the Blowing Stone has been moved 

 to its present position. Nor can I find any record of Sarsens on the 

 escarpment. The highest point at which they occur is one mile from 

 the edge of the Chalk escarpment. Mr. W. Cunnington, who knows 

 the Wiltshire and Berkshire Downs better than anyone, says that their 

 absence from the ridge is certainly not due to their having been 

 removed by man ; and my colleague, Mr. Andrews, has recently 

 searched the downs around Aldbourne and can find none nearer the 

 escarpment than on Kingstone Down. Their absence from the 

 escarpment seems to show that the Sarsen-yielding horizon thinned 

 off against the higher part of the Chalk, so that this apparently rose 

 above the level of the water in which these sands were accumulating. 

 The estuary probably extended up the Kennet Valley and then over 

 the lower slopes of the Chalk Downs to Abury, and north of this till it 

 reached the Jurassic plain about Swindon. A further continuation 

 may be indicated by some Sarsens which occur in the valley that 

 runs from Swindon to Hannington and near Highworth. But in any 

 case the Sarsens in the Swindon district are sufficient to prove that 

 at least a part of the Berkshire Chalk escarpment was in existence in 

 the time of the later set of Sarsen-yielding sands, i.e., probably Upper 

 Bagshot. 



That the Berkshire escarpment took up its present position earlier 

 than that of the Chilterns is shown very clearly by a comparison of 

 their structure. (Figs, i and 2.) Thus, while the Chilterns is jagged 

 and irregular and breached by numerous former stream courses, the 

 Berkshire escarpment is regular and unbroken. The whole drainage 

 system of the Chilterns has been altered since the date when the 

 escarpment was formed, while in the Berkshire Downs the drainage is 

 entirely resultant on the present slope of the country ; thus the 

 drainage of the latter is of the type known as subsequent, that of 

 the former as antecedent. 



There is a third line of argument which there is only space 

 here briefly to indicate, viz., that the Thames and its tributaries 



