RECENT ADVANCES IN SCIENCE 363 



This is due to the low temperature generally found at the surface 

 at about 8 a.m., when most of the observations were made. 

 The topographical factor further complicates matters, e.g. at 

 Mount Weather in January it is the westerly winds which are 

 the coldest, and the easterly the warmest, because of the near 

 neighbourhood of the Atlantic towards the east. 



Owing to these various causes, there is no simple relation- 

 ship between the temperature differences and the wind, and it 

 was found necessary to make a diagram for each station, having 

 the wind direction and time of year as ordinates and abscissae 

 respectively, on which isopleths (lines of equal temperature 

 difference) could be drawn. As was only to be expected, the 

 form of the isopleths was found to be very different for each 

 station. Similar diagrams were made for a height of 1,000 

 metres. The method of use is to find the point on the diagram 

 corresponding to the given time of year and direction of the 

 wind, and then observe which isopleth is nearest to this point. 

 The value of this isopleth must then be subtracted from the 

 surface temperature, so as to obtain the best estimate of the 

 mean temperature required. 



The author goes rather fully into the question as to how 

 accurate these estimates are, and those of the pressure at the 

 top of the column calculated from them. He uses modern 

 statistical methods to determine the probable error, and finds 

 this to be the order of i or 2 millibars. From this he hopes for 

 good results from the method when free-air observations are 

 extended to a larger number of stations. 



Winds and Temperature Gradients in the Stratosphere. — 

 G. M. B. Dobson, in the Quarterly Journal of the Royal Met. 

 Soc. (January 1920, vol. xlvi. No. 193), discusses the winds of the 

 stratosphere over North- West Europe. The well-established fact 

 that in the neighbourhood of the British Isles the troposphere 

 is colder in cyclones than in anticyclones, whereas the strato- 

 sphere is warmer, implies that the wind velocity should diminish 

 above the tropopause, as has often been pointed out. The 

 results of seventy ballons-sondes ascents which penetrated well 

 into the stratosphere have been dealt with in a rather novel 

 manner. In addition to finding average values of wind velocity 

 for different heights above the ground, diagrams were also 

 made in which|the height of the tropopause was taken as the 

 zero line, and the velocities were plotted for various heights 

 above and below this line. Moreover, the ascents were separ- 

 ated according to the velocity of the wind, light and strong 

 winds being dealt with separately. An interesting fact immedi- 

 ately became evident, namely that at some height between 

 8 and 14 kilometres nearly all winds above 15 metres/sec. fall 

 off to under 10 metres/sec. This takes place very close to the 



