138 Mr Lawson on the Trade- Winds at Barbadoes. 



12. It has been stated above, that the temperature of the 

 sea at noon, at Barbadoes, in April, was 79°-3 Fahr. (6) ; and, 

 that in September, that of the current which sets past the 

 island was 81°.7 (7)- No error of any consequence will be com- 

 mitted, if the mean of these 80°.5 be assumed as its tempera- 

 ture round Barbadoes during May and June. I have elsewhere 

 shewn, that the mean temperatures of the air in the latter 

 part of May and June 1841, there, were 79''-77 and 80°. 40 

 Fahr., and the mean maxima 83°.8G and 85°.23 respectively;* 

 the former almost equal to the temperature of the sea in the 

 vicinity at noon, the latter exceeding it, and, consequently, that 

 of the superincumbent air (8). This unequal distribution of 

 temperature in the air, caused its ascent over the land during 

 the day, while the cooler air flowed in from the sea to supply 

 its place, at the same time that the regular trade-wind con- 

 tinued at a moderate elevation, as shewn by the drift of the 

 lower clouds. These observations having been made at the 

 south<west part of the island, the wind naturally drew to the 

 south-ward during the day (it was frequently from ESE., 

 while the lower cumular clouds were drifting from E. or ENE.) ; 

 and on one occasion, while it had been the greater part of the 

 day at SSE., the lower cumuli were drifting with consider- 

 able rapidity from E. I do not mean, however, to state, that 

 all the north-easterly winds during May and June were of the 

 kind just described, for, in a good many instances, the lower 

 clouds moved in the same direction as the wind. 



13. In July, the mean temperature had fallen to 80°.05 Fahr., 

 and the mean maximum to 84°.72,t while the temperature of 

 the sea must have increased. The excess of temperature in 



* Edlnburgli New Philosophical Journal, June 18i3. 



t As maximum and minimum thermometers give the extremes of tem- 

 perature during a day, without reference to the hour at which these 

 occur, their indications for a month must always be, the former higher, 

 and the latter lower, Ihan the corresponding points in the mean diurnal 

 curve of temperature, and these differences will be greater as the wea- 

 ther during the period was more unsettled. The weather at Barbadoes, 

 in July, was particularly unsettled ; frequent heavy showers, accom- 

 panied by sharp squalls, occurring during the day, which always caused 

 a considerable reduction of temperature. It is not a little remarkable, 



