SPRAYING PROBLEMS 155 



town, Dumfries, and Ayr in July, and there is a 

 record which gives the first week in August for the 

 Island of Islay (Hebrides). The records for Argyll, 

 Linlithgow, Renfrew, and Fife counties to the north 

 or north-east are not earlier than the middle or end 

 of August. In Perthshire, farther north, and Had- 

 dington, to the east, disease is not recorded earlier 

 than September, whilst it is stated to be unknown 

 in Inverness, Aberdeen, Elgin and Moray, and 

 Caithness. 



The course taken by the epidemic in England is 

 perhaps determined to a certain extent by the some- 

 what zigzag valleys of the Wye, Severn, Trent, 

 Welland, Great Ouse, and Thames, which radiate 

 in fan-wise direction towards the east, north-east, 

 and north when viewed from the Severn estuary. 

 These valleys, subject to mists and fog, would prove 

 ideal channels along which the disease, helped by 

 prevailing southerly and south-westerly winds, could 

 spread over the country-side. 



It has to be remembered that the direction of 

 development east and west also corresponds to pro- 

 gressive seasonal lateness. The fact that there is 

 an interval between the attack on early and late 

 varieties seems to indicate a period in the life-cycle 

 when the plant is more susceptible to the disease. 

 If this were the case disease would appear progres- 

 sively later from the south - west and west to the 



tf 



eastern and northern counties. 



If an outbreak of disease occurs in the present year, 

 it may be expected to take the same general course ; 

 but its date of appearance and degree of severity may 

 vary considerably. For example, in 1917 disease 

 was comparatively light in the counties of Cheshire 

 and Lancashire, and did not make its appearance 

 until the end of the third week in August; in 1845 



