The maps of August 27 and 28 show marked deterioration of the low index 

 characteristics of the general circulation north of latitude 50° N. Note the continued 

 slow genesis of the Icelandic low, the disappearance of the northern European anti- 

 cyclonic cell as well as the absence of any important polar anticyclone on either 

 map, and finally the first appearance of a significant cyclonic center in northwestern 

 Canada. South of latitude 50° N the circulation pattern remains very weak. 



It appears on the map of the 28th that a new tropical air fog condition is 

 being established along the Atlantic coast from Long Island to Newfoundland. 

 Contrary to the conditions in winter (see p. 50) the occurrence of tropical air fog at 

 inland locations in summer is relatively infrequent and of small significance. 



An interesting feature of the map of the 28th is the indication of a tropical 

 disturbance located near eastern Cuba. Minor undulations of the isobars in the 

 Caribbean area have been apparent several times during the preceding week, but 

 this is the first such occurrence where the winds have been sufiiciently accelerated 

 to give clear evidence of a cyclonic center. The genesis of tropical disturbances in 

 the vicinity of the Sargasso Sea, Caribbean and the Gulf of Mexico is favored by 

 weakness of the Atlantic anticyclone in the region where the formation occurs. 

 Some horizontal convergence seems to be required before the cyclogenesis can begin. 



Between August 29 and August 31 the trend towards a stronger circulation 

 pattern continues steadily. Both the Icelandic low and the cyclonic cell over west- 

 ern Canada become slowly larger and stronger, thus intensifying the zonal wester- 

 lies south of the centers. Significant anticyclonic centers at the higher latitudes 

 remain conspicuously absent, while a strong Azores high develops in the eastern 

 Atlantic. For the first time during this map series, a cyclonic center crosses from 

 extreme northeastern Canada to southern Greenland and merges with the Icelandic 

 low. It is this development, which is so characteristic of a strong circulation pat- 

 tern, that leads to the intensification of the Icelandic center between August 31 

 and September 1 (cf. conditions from January 13 to January 14, p. 48). 



It is interesting to note in this case how the increasing strength of the 

 zonal westerlies in the vicinity of latitude 50° N carries the warm foggy maritime 

 tropical air from the Grand Banks region rapidly eastward across the North Atlantic. 

 The fog and drizzle conditions which characterize this warm air mass (more drizzle 

 and less fog as the wind velocity increases) reach the British Isles on September 1. 

 In the preceding case of widespread fog formation in the same Grand Banks fog 

 center (August 20) a typical low index circulation pattern carried the warm foggy 

 air mass northward to Greenland (August 21-23). It is unusual to get a combi- 

 nation of conditions favorable to widespread fog formation (low index pattern) and 

 rapid eastward transport (strong circulation pattern) in such a manner that New- 

 foundland fog reaches the British Isles. 



Only in the southwestern Atlantic does the circulation pattern between 

 August 29 and September 1 manifest a tendency towards weaker rather than 

 stronger characteristics. These characteristics are seen in the slow northwestward 

 drift of the tropical disturbance which is moving through the eastern Gulf of 

 Mexico, and the almost stationary condition of the second tropical disturbance 

 which developed north of Porto Rico on August 30 and 31. The slow movement of 

 the first storm even after it starts to recurve northward, and the almost motionless 

 state of. the second storm testify to the absence of any predominant wind move- 

 ment in this area even as late as September 3. The second disturbance later recurved 

 off the east coast of Florida and moved very slowly northeastward parallel to the 

 coast. It attained full hurricane intensity near the center before it reached 30° N, 

 and apparently retained hurricane intensity at the center as late as September 11 



54 



