Captain IlKi'WORTH has railed attention to the comparative warmth of southerly, i.e. polar, winds, which 

 was also a special feature of the observations at (.'ape Ad.-u-e liy the "Southern Cross" expedition. 

 Mr. 1 JINKS lias dealt with the extraordinarily low summer temperatures in relation to the dryness of the 

 atmosphere : lie lias also referred to the locally high temperature of the station at the Winter Quarters of 

 the " Discovery," and has suggested an explanation. Dr. CHREE has dealt with the diurnal inequalities 

 and the extremes of temperature. There are other points which may be noted. The great range of 

 temperature during the winter months is very remarkable; the range in June, July, and August, 1903, 

 during the Antarctic- night was 64 F., 66 '2 F., and 65 F., respectively, and closely connected with this 

 point is the extraordinarily rapid fluctuation of temperature as shown by the thermograph. A specimen 

 of the thermograph traces, giving examples of these rapid fluctuations of temperature, is reproduced here 

 (fig. 1, p. xi.). 



There is additional evidence of the existence of such rapid fluctuations. In default of any instructions to 

 the contrary, an alcohol minimum thermometer, mounted in the screen beside the mercury thermometers, 

 was read and set every two hours. Mr. CURTIS has called my attention to the fact that the minimum 

 reading for the interval is in nearly every case below the mercury readings at the beginning and end. 



Such would certainly not occur in any ordinary climate. So persistent is this occurrence of a minimum 

 every two hours that Mr. CURTIS is disposed to doubt the accuracy of the readings as not being always 

 borne out by the thermograph curves, but if there be an error of reading it is difficult to say which of the 

 two thermometers is liable to the error, and in very many eases a depression of temperature within the 

 two-hour interval is shown on the thermograms. 



In connexion with these extraordinary results for temperature it must be remembered that when the 

 physical process going on at the surface is a transference of heat from the air to the ground, cooled by 

 radiation or otherwise, very capricious results are likely, because the cooled air remains at the surface. 

 In any locality the fluctuations of minimum temperature on a clear night are, for that reason, far greater 

 than the fluctuations of maximum temperature on a cloudless day. This property of cooled air must be 

 borne in mind in considering all questions of Antarctic surface temperatures. The coldest temperatures 

 are likely to be found in stagnant air, and any wind that disturbs stagnant air will be a warm wind. 

 Kvery fluctuation in the strength of the wind will, similarly, cause an alteration of the conditions 

 governing the temperature. From this point of view it is to be regretted that the variation of 

 temperature with height, within such moderate compass as can be reached on the ship, up to 50 or 

 60 feet, was not made the subject of experiment or observation at comparatively short horizontal 

 distances apart. 



On p. 429, and elsewhere, Captain HEPWORTH suggests that the experience of relatively warm winds 

 from the polar quarter may be attributed to the circulation of warm air in cyclonic systems, by which the 

 current brought to any point from the southward may be traced to some equatorial quarter. It is difficult 

 to know how far such an explanation may be carried. Without some limitation it would enable one to 

 assign any desired characteristic as regards temperature to any part of the storm area. The detailed 

 work upon the paths of air in travelling storms of temperate northern latitudes in " The Life-History of 

 Surface Air-Currents " tends to show that it is difficult to give any general rule for the ultimate origin of 

 the air supply for any particular part of a depression. In our experience in these latitudes, winds from 

 equatorial quarters are generally warmer than those from polar quarters. But there is so little information 

 about the mechanism of a travelling depression in the Antarctic regions that no final opinion is possible 

 with regard to the question of the association of temperature and wind direction. 



I'l-ivxuri' iiinl ll'iiul*. 



The question of the prevailing wind at Winter Quarters and its relation to the distribution of pressure 

 is another interesting point which is not entirely disposed of. The idea of an anticyclone, possibly of no 

 great vertical height, but spreading over the surface of the South Polar continent, has found expression in 

 many forms in meteorological literature; the references to HANX and SUPAN given by SCOTT in the 



