97 



In the cuttings for the North British Railway, between Arthur 

 Seat and Musselburgh, the upper sides of the largo boulders are ge- 

 nerally found smoothed and scratched. The scratches seem to be from 

 NW. to WNW. by compass. On some of the boulders there are in- 

 dications of more recent scratches running W. ^ S. by compass. 



The boulders in the railway cuttings between Haddington and 

 Dunbar exhibit scratches running from NW. to WNW. 



The opinion formed by the author on these data was, — 



(1.) That the agent which had polished and scratched the rocks 

 on Arthur Seat, was the same as that which had polished and 

 scratched the boulders. 



(2.) That it had acted from the north-westward over a large and 

 low district of country. 



(3.) That the polishing and scratching had been effected by the 

 gravel and angular blocks existing in the boulder clay and diluvial 

 gravel. 



(4.) That there had been rushes of water along the country, 

 which bore along the mud, sand, gravel, and boulders now spread 

 over the country, and which, in passing over the rocks and large 

 boulders, smoothed and rutted them. 



(5.) That, at this period and subsequently, water must have stood, 

 in a comparatively tranquil state, above the level of Sampson's Ribs, 

 to account for the beds of sand existing on the south side of Arthur 

 Seat, and at a level of 200 feet above Duddingston Loch. 



(6.) That the outline or configuration of the district, thus sub- 

 merged, could not have been materially different from what it now 

 presents. 



3. Results of the Makerstoun Observations, No. II. On the 

 Relation of the Variations of the Vertical Component 

 of the Earth's Magnetic Intensity to the Solar and Lu- 

 nar Periods. By J. Allan Broun, Esq. Communicated 

 by General Sir T. M. Brisbane, Bart. 



The following results are deduced from the observations of the 

 balance or vertical force magnetometer described in the Introduction 

 to the Makerstoun Observations. The observations were con-ccted 

 for temperature by a method previously described. 



The diurnal variation of the vertical component consisted, in the 

 year 1844, of three maxima and three minima, occurring as fol- 

 lows : — 



VOL. II. K 



