Clare Island Survey — Climatology. 6 3 



In the analysis of the lighthouse records we have considered wind 

 direction from only eight points ; and we have divided the intensities into 

 three grades, viz. : — 



Beaufort number. Description. Velocity in miles 



per hour. 

 to 3 inclusive. Calm to gentle breeze . to 12 . 



4 ,, 7 ,, Moderate breeze to high wind. 13 to 38 



8 „ 12 ,, Gale to hurricane. Above 39 



With those broad divisions it is reasonable to regard the observations as 

 quite satisfactory for our purpose. 



The study of the wind records was conducted with the following objects 

 in view : — 



(a) To determine for each month the relative frequencies of the winds 



from the eight main points. 



(b) To determine for each month the relative frequencies of each grade of 



intensity of the wind in each of the eight directions. 



(c) To consider in general and for each month whether winds of any 



particular intensity (i.e. light, moderate, or strong) were more 

 frequent by day (8 a.m., noon, 4 p.m.) than by night (8 p.m., 

 midnight, 4 a.m.). 



(d) To consider whether in passing from day to night there was evidence 



of any general tendency for winds to change in intensity or in 



direction. 

 The reduction of the observations was effected in the following way. 

 The six records for each day of each month were transferred from the 

 weather journals, and entered so as to distinguish between day and night 

 observations, and to classify the intensity in one of the three grades already 

 specified. The records were in the first place referred to the sixteen points 

 as regards direction ; then the number of occurrences of wind from an 

 intermediate point — e.g. ENE. — was divided equally between the two corre- 

 sponding main points NE. and E. In attempting to reduce the occurrence- 

 frequencies to percentages in a uniform way, it was noted that the four 

 months with thirty-one days gave 6x31x17 or 3162 records in all, whereas 

 the months of thirty days gave only 6 x 30 x 17 or 3060 records. The 

 numbers for these latter months were in each case increased in the ratio of 

 3162 to 3060, so as to bring all the months to a common value for 

 comparison. 



A2 



