354 



SCIENCE PROGRESS 



morning hours are specially suitable for flying, because they 

 make starting and landing easy and these are the most difficult 

 periods of a flight. The explanation usually given of the 

 meteorological fact that the wind falls off at night, which is 

 quite well known, is that during the day the lowest layers ol 

 the atmosphere are constantly churned up by the effect of the 

 sun upon the surface of the ground and that at night, on the 

 contrary, the cooled surface air can remain on the ground undis- 

 turbed by convection currents. Thus the effect of obstacles on 

 the ground and surface friction is much more marked. The 

 surface layer may be actually at rest while the motion goes on 



-Bureau Central /Vle^brokjgique- 



8 Nixal 4 8 



— Liiffe! Tower 



8 N<"-n 4 O Mdt 





Means for the six years 1890-1895. 



Fig. 4. — Isopleths of wind velocity for different hours of the day in different months of the 

 year at the Bureau Central Meteorologique, Paris, and at the top of the Eiffel Tower 

 (1,000 ft. above ground). The isopleths are drawn for steps of a metre per second. The 

 lines on the left-hand diagram are for 1, 2 and 3 metres per second, those on the right- 

 hand diagram for 6 metres per second and successive steps up to II metres per second. 



overhead even more vigorously than in the daytime, because the 

 influence of the surface does not reach so far as in the day. 

 Presumably the night curve of variation with height is a modi- 

 fication of the straight line to a curve something of the form 

 represented in fig. 5 ; we have no observations from which to 

 deduce its actual shape. 



Balloons and airships manoeuvre mostly within 3,000 ft. and 

 it is the gradient wind with which they have to deal except in 

 starting and alighting. In computing the path of a free run the 

 gradient gives better information than the anemometer. 



