a04 ROYAL SOCIETY OF CANADA 



crease, but instead would call for a more or less abrupt beginning, which 

 is not the case. Inteirmittent rhythmic vibrations of the ground, nearly 

 synchronizing with the period of the pendulum, setting the pendulum 

 in motion, could produce the phenomenon. 



'The validity of a supposed relationship between different pheno- 

 mena, as cause and effect, is readily tried by predicting the effect when 

 given the cause. 'This has been done with reference to the existence of 

 a " Low " in the Gulf and a " High " over the Atlantic coast to the south, 

 or in general by taking the daily weather map with its isobars and from 

 it predicting the resulting microseisms. The result has in so far been 

 satisfactory that in the large majority of cases the microseisms have 

 fairly well answered in presence and magnitude the prediction. There 

 are, however, still important outstanding differences that require further 

 explanation. Just why the " Low " about the Gulf should have such an 

 influence in the production of microseisms is by no means apparent. The 

 two main physical features are the shallow Gulf and the St. Lawrence 

 valley in which lies the Great St. Lawrence and Champlain Fault, 700 

 miles long, already referred to. As secondary, is the general trend of 

 the Atlantic coast, and possibly that too of the Alleghany mountains. 



On frequent occasions there is a " Low " over the Gulf, another 

 " Low " over Arkansas, while one " High " rests north of Lake Superior 

 and another over Bermuda. When those conditions obtain with steep 

 gradients we are pretty sure to have marked microseisms. 'The line of 

 the " Lows," then, lies in the St. Lawrence valley, while that of the 

 " Highs " is at right angles to the former. In this case the maximum 

 strain is along the valley of the St. Lawrence, along the Great Fault, so 

 that from a priori reasoning marked microseisms might be expected. 



In concluding the present investigation of the well-marked micro- 

 seisms recorded here, we will repeat that the presence of a " Low " over 

 the Gulf surrounded by steep or fairly steep gradients on a given morn- 

 ing is indicative of more or less well-marked microseisms following at 

 Ottawa that day. 



It has already been stated that the large majority of microseisms 

 have a period of about 5s with small fluctuations. The reason for the 

 fluctuations is by no means apparent unless it be the varying depth of 

 the earth's surface involved. 



On occasions the above common period changes to one of about one- 

 half, or about 3^, showing, however, a transition time during which 

 there is an irregularity and interference, so that their period is unre- 

 cognizable. At present no explanation can be offered for this sudden 

 change. When the period is so short the amplitudes are very minute, 

 although visible to the naked eye. 



