514 



evidence of a diurnal variation in the summer months. The results are given in Table II., and graphically 

 in fig. 3. They show that the mean duration ]KT hour is very uniform between hours and 3 a.m., from 

 which hour it increases steadily till noon; then it again remains without change for about 3 boms and 

 steadily diminishes till 11 p.m., which is the least sunny hour of the twenty-four. The total range of the 

 diurnal variation as shown by these monthly totals amounts to 4 1 hours. 



TAISI.E II. Diurnal Variation of Sunshine in Summer. 

 (November, December, January, 1902-4.) 



Hours. 



1. 2. S. I. 



6. | 7. 8. !. 10. 



11. Noon. 



1. 2. 



5. 8. 



10. 11. 



Mi.lt 



(Average monthly .luralion (in hours) of sunshine for each hour of the day.) 



12-1 12-0 12-0 I 12-2 | 13-8 13'4 13'7 ll'l irii|l.Vl 1.V5 j l.Vti 15 '5 15'5 15'5 15'3 ld'l | 14't) 1-1-4 IS'-I | 12 I ll'ti 11 ' 12'n 

 I I 



Fig. 3. 



SOLAR RADIATION. 



The record of the Campbell-Stokes' burning sunshine recorder is primarily one of duration, and the 

 information it affords respecting the intensity of the solar radiation is only of a very general character, and 

 is obtained from the character of the burn made on the card. A marked feature of the Antarctic records 

 is the sharpness of the burn, which generally goes right through the card, and in not a few instances, 

 whore the record has continued throughout the entire day, it has severed the card longitudinally in two. 

 This indicates a particularly strong solar action such as could have been expected only in an exceptionally 

 dry and dust-free atmosphere. 



The most direct measurement available of the intensity of the solar radiation is that afforded by the 

 black-bulb thermometer /';/ /w/w, and the readings of this instrument made at "Winter Quarters are very 

 remarkable. The observations made daily throughout the four months November to February, during 

 nearly the whole of which time the sun was continuously above the horizon, have been measured, for each 

 month separately, and the results are exhibited in Table III. 



For purposes of comparison the results of similar observations made at Madras during the four summer 

 months of May, June, July and August, 1890, are also given in the table, the year 1890 having been 

 chosen only because it happened to be the one most easily available. It is obvious that the full significance 

 of the figures given can only be appreciated when the difference between the sun's meridian altitude at the 

 two places is taken into account. 



At Discovery Bay, during the months of November and Fein-nary, the daily change in the height of the 

 sun above the horizon was considerable, but even when it had reached its highest altitude in December it 

 was still more than 60 from the zenith. At Madras, on the other hand, a nearly vertical noon sun was 

 experienced throughout the four months quoted. Yet under these very different conditions the mean solar 

 temperatures ;it. Di.-eovery Day in December and January were only 1-i" less than the similar means at 

 Madras in June ami July, and the maximum reading at Discovery Bay in December approached to within 

 .1 nf the June maximum at Madras. 



