PHYSICS 



similarity and uniformity of all except the surface curve 

 are remarkable. We cannot build much theory on such 

 curves, as the curves of ice-temperature represent single 

 observations in each month, while the air curve is a real 

 mean deduced from twenty-eight daily means. Still the 

 similar course traced by each curve cannot be entirely 

 chance. If there were no direct relation between the air 

 curve and the others we would expect greater differences 

 at different depths. 



The monthly mean is selected as giving the nearest 

 estimate as to the rate at which the temperatures within 

 the ice follow the air temperatures. The daily mean of 

 air temperature would not be expected to correspond at 

 all with the slower changes within the ice, but the curve 

 was drawn, and fluctuated extremely. Then the mean 

 for a week before each observation was taken, also the 

 weekly means for the entire period, and they showed no 

 obvious relation to the ice-temperatures. Even with the 

 means for a fortnight before each monthly series the 

 correspondence is far from close. 



When the curve of the monthly means is drawn beside 

 the others it is at once evident that it takes a place in the 

 series, but that its range is much greater. All the curves 

 converge steadily after August, and approximate very 

 closely in November and December. 



MINERALS AND CHEMISTRY 



Notes by DOUGLAS MAWSON, B.Sc, B.E. 



A large variety of minerals, chiefly rock-forming 

 types, were met with by the expedition. The minerals 

 included felspar, pyroxenes, amphiboles, micas, garnet, 

 &c. Among the most notable features was the occur- 

 rence of idiomorphic felspar crystals a couple of inches in 



365 



