CLIMATE. 361 



As is usual in these regions of the Pacific, the fluctuation of the 

 barometer, whether daily, yearly, or monthly, is very small. Thus, 

 the range during the 22 months we passed in the group, was from 

 2983 to 30-18 inches, or about a third of an inch ; whilst the 

 average monthly range was rather under a quarter of an inch, and 

 the usual diurnal variation about 'O^ of an inch. 



Whilst endeavouring to compare the temperatures of the different 

 seasons, I have mainly used the Ugi register, since it gives a con- 

 tinuous record for more than a year. At Ugi in 1883, the 

 portion of the year from June to September inclusive was slightly 

 the coolest, but the difference in the means was not 2° ; and, in 

 truth, taking all the thermometric observations into consideration, 

 the seasons are scarcely distinguished by their temperatures. As 

 Lieutenant Leeper^ remarks in his report, the temperature varies 

 but little all the year round, the monthly mean varying between 

 80° and 8.5°. The annual mean temperature may be placed at 82° 

 to 83°, and the range from 75° to 95°. The daily variation is 

 considerably affected by the exposed or protected position of any 

 locality at the coast. Judging, however, from the data at my 

 disposal, it is usually less than 10°, e.g., 79° at night, and 88° at 

 mid-day. 



From the hygrometrical observations, it may be inferred that 

 the climate of these islands is generally very moist. The relative 

 humidity, taking 100 as saturation, ranged at Ugi in 1883 from 54 

 to 100 ; but the monthly range was usually from 72 to 95, the mean 

 for the year being 83.^ This mean degree of relative humidity is 

 much greater than that of Levuka in Fiji which would seem to 

 average about 70 f but in truth tliere is little necessity for me to 

 remai'k further on this well-known feature of the climate of these 

 islands. Yet, I should add that this proportion of aqueous vapour 

 would not necessarily be oppressive in a temperate latitude. In a 

 tropical climate, however, any influence that retards the evaporation 

 from the skin of the normally excessive perspiration, is a cause of 

 personal discomfort, such as would not be experienced in a drier 



^Vide Quart. Journ. Eoy. Met. Soc. vol. XL, p. 309. The instruments used on board 

 were previously verified at Kew. From want of leisure, Lieut. Leeper was unable to do 

 much more than tabulate bis observations. I have therefore extracted from them such 

 general facts and inferences as they sustain. 



-There are no observations for January, but since the mean relative humidity varies with 

 the rainfall, I have approximately estimated that for January to be 83. 

 2Iieut. Lake's observations for 1876 and 1877. (Quart. Journ. 31et. Soc.) 



