THE HEART OF THE ANTARCTIC 



was warmer. The two curves diverge in April, meet 

 in May and June, diverge in July, meet in August, 

 diverge in September, meet and cross in October. Though 

 there is only one crossing the chain is almost as distinct 

 as in the Scotia records. 



Our curve for 1908 is first noticeable for the striking 

 parallelism from March till August, with the Discovery 

 second year. It lies between the two Discovery curves in 

 March and April, rises greatly in May and equals 

 Borschgrevinck's. In June it is well above Borschgrev- 

 inck's, making the least cold June known in Victoria 

 Land. From August to November it is again the warmest 

 season known in this region. 



In all the records for Victoria Land the trough 

 form of the curves is very noticeable. The summer, as 

 measured by temperature, is very short : the approximate 

 maximum is maintained only for two months (December 

 and January), and the curve falls away steeply and 

 steadily on each side. The temperature falls rapidly till 

 the approximate winter mean is reached, after which it 

 fluctuates about this mean for from four to six months, 

 before the next summer's rise sets in. 



The winter mean was reached approximately as early 

 as April in the second year of the Discovery (1903), and 

 in 1908. In the Discovery first year (1902) it was reached 

 in May, and in 1899 (Borschgrevinck) as late as June. 

 The rise began in all cases in September, except in the 

 second year of the Discovery (1903) , when it was a month 

 later. 



In Dr. Bruce's records the summer is longer, lasting 

 for five or six months, and the winter is much less marked, 

 as far as temperature goes. The difference between the 

 highest monthly mean for summer and the lowest for 

 winter is only about 26° (Fahr.) . In 1908 the difference 

 was 47°, 



416 



