332 TRANSACTIONS OP SECTION A. 



in isochronous lines indicating the advance of the 'crochet' with its attendant 

 squall with a linear front, which passed Scilly at about 2 p.m. on July 29th. 

 Over England the front was generally ranged from N.W. to S.E. and it ad- 

 vanced with the S.W. wind, but the lines were bent apparently over the Irish 

 sea and the direction of the front over Ireland was not well traced. The dis- 

 turbance advancing with this front was attended nearly everywhere by thunder- 

 storms, but in the north of England its intensity was much diminished and it 

 was barely recognisable in barograms for the north of Scotland on Sunday morn- 

 ing. The anemograms showed that at many stations in the south the wind 

 squall commenced at the full strength of gale force without any preliminary 

 gusts. 



The peculiarity of the disturbance is its M shape and its line of advance in 

 front of a S.W. wind. It is thus distinguished from the ordinary type of line 

 squall, which has a V-shaped 'crochet' and comes from the W. or N.W. 



The ordinary line squall is explained by the setting in of a cold current of 

 westerly air which suddenly replaces a warmer current from S.W. or S. A 

 similar explanation probably holds for the cases now under discussion, but the 

 circumstances of the instability in the atmosphere which gave rise to the pro- 

 gressive thunderstorms are not so clear. The records of temperature, wind 

 direction, and rainfall have still to be examined with the object of tracing the 

 physical processes represented. 



2. Report of the Seismological Committee. — See Reports, p. 30. 



3. Note on the Periodogram of Earthquake Frequency from Seven Years to 

 Twenty Years. By Professor H. H. Turner, F.R.S. 



1. The Catalogue of Large Earthquakes recently edited by Professor Milne makes 

 it possible to examine the periodogram. The present note is restricted to the portion 

 from seven years to twenty years. The methods of Professor Schuster for exhibiting 

 the results of such an examination are still not very widely known, and are more- 

 over capable of a variety of forms. Hence, enough of detail is given here to enable 

 others to use the method if they wish. 



2. Let us first consider the period seven years. To examine whether there is a 

 seven-year periodicity we first arrange the yearly numbers in groups of seven 

 years, and it is convenient to take five groups together, thus : — 



Earthquakes recorded in 1700-1734. 



The sums for the consecutive groups are first multiplied by sin 6 and then by cos 6 ; 

 putting fl successively equally to 0°, 51°, 103°, 154°, 206°, 257°, 309° (multiples of 

 360°/7). Now write 



a = -19x200/167 6 = -14x200/167 tan 1 a/&=234° 



= -22-8 =-16-8 a 2 +& 2 =802 



The multiplication of a and b by the constant 200, and division by 167, the sum 

 of all the numbers, is a comparatively new suggestion arising out of the present case, 

 but applicable to other cases. When we record rainfall or sun-spots the average does 

 not change seriously ; but in the case of these earthquake records facilities of com- 

 munication have changed it very greatly. Compare, for instance, the latest group 



