1907.] 



M ktk< )K( )i;of;i< a i . Statists :s. 



13 



light " falls below zero, showing that the chance of the week 

 being without rain is greater than one in twelve. This feature 

 is found to occur in the statistics for aU districts ; it is par- 

 ticularly marked in the Nortli of Scotland. 



Another interesting point is the large divergence from the 

 average of the limit for "very heavy" rainfall about the 

 middle of autumn in District 5, showing that weeks of heavy 

 rainfall arc comparatively frequent in that season in the South 

 of England. This feature is shared by District 8, " England, 

 South-west," but it is much less marked farther to the Nortli. 



In the sunshine curve for District 5, the most noteworthy 

 feature is the relative scantiness of sunshine about the 24th 

 and 30th weeks — second week in June and last week in July. 

 The deficiency in the 24th week finds its counterpart in a slight 

 inflection of the curve for accumulated temperature above 42 0 . 



All the temperature readings used in computing the tables 

 of accumulated temperature in the Weekly Weather Report 

 are from thermometers exposed in screens of standard pattern 

 at a height of about four feet above the ground. It not in- 

 frequently happens on calm, clear nights that the temperature 

 on the ground falls several degrees below that in the screen, 

 and sharp ground frosts which do considerable damage to low 

 growing vegetation may occur without the screen temperature 

 falling far below the freezing point. The incidence of ground 

 frosts is often very local and depends very largely on the topo- 

 graphy of the land, and thus it is not practicable to combine 

 the readings from thermometers exposed on the grass at a 

 number of stations into mean values for districts. Information 

 as to the incidence of ground frosts is given for a number of 

 stations 111 .1 separate table in the Weekly Weather Report. 

 The data are summarised by stating the lowest temperature 

 dtt the grass experienced during the week and the number of 

 occasions on which the thermometer fell below the freezing 

 point by at least 2\ For several of these stations the mean 

 temperatures of the soil at depths ol one foot and tour feet are 



also given. 



R. G, K. Lempfert, 



