April 4, 191 1 



NATURE 



87 



p7ooo ft., with a slight rise of temperature above 

 before it was disturbed by the cumuli rising 



)m below. 



On days when the disturbance is due mainly to 

 ■heating at the surface,' and no cloud-sheet or 

 damp layer exists at the height of a few thousand 



I 



• Fig. 2.— Strato-cumulus clouds. Tops 7000 ft. to 8000 ft. Ri-.eof tentperature 6° F. in 

 100 ft. .nbove highest parts. Taken, from about 8000 ft., Brooklands, ii' a.m., 

 August 16, 1917. 



lower air is very damp, showers may fall from a 

 much lower level. Fig. 3 shows tops of cumuli 

 at about 9500 ft. joined to a shower on the right, 

 and also patches of stratus. 



We have hitherto only been considering clouds 

 which are due to disturbances originating at the 

 surface. The majority of the upper clouds 

 are entirely independent of superficial dis- 

 turbances. The same is true of many 

 types of rain-cloud, which are due to 

 upward movement on a large scale ; even 

 thunderstorms sometimes develop with 

 their bases at the height of fully 8000 ft., 

 and are indep)endent of disturbances ori- 

 ginating at the surface. In many cases 

 rainfall develops over a wide area from 

 high cloud-sheets, which gradually extend 

 downwards as the upward movement be- 

 comes more pronounced. The rain area 

 may advance over a wide tract of coun- 

 try, preceded by a high cloud-sheet. The 

 rain-clouds consist of thin mist extending 

 to a great height, and in winter, and at 

 any season above 10,000 ft., they may 

 consist only of thin snow. In France the 

 snow-storm of January g, igi8, developed 

 from a cloud-sheet of this type, which ori- 

 feet, the clouds usually retain the form of cumuli, ] ginally came over high up, and later at lower 

 which are, as a rule, contained within a definite ' heights, the snow being finally continuous from 

 layer the level of which rises during the day. | the ground to a great height. Rain-clouds of this 

 W'hen the clouds have risen to 2000 ft. or more \ type may be uniform over wide areas, and there 

 above the surface, the lapse-rate near the ground ! may also be turbulent cloud-sheets near the sur- 



is adiabatic, but 



this is not usually 

 the case at the 

 cloud level, with 

 the result that the 

 clouds at the top 

 of the ascending 

 currents are usu- 

 ally colder, than 

 the surrounding 

 air, being forced 

 ".p by small irre- 

 gularities of prc^ 

 -lire. Once th( 

 lapse-rate near 

 the ground is 

 adiabatic, large 

 bodies of air may 

 be forced up in 

 this way, form- 

 ] n g banks of 

 cumulus clouds. 

 If these enter a 

 aver the lapse- 

 ate of which is 

 ibove the adia- 

 atic for saturated 



ii, these large cumuli may become warmer 

 iian the surrounding air and continue to 

 scend, and perhaps finally cause showers or 

 ninderstorms. The tops of thunderstorms are 

 I 



C. K. M. D. 

 Fig. 3.— Cumuli, with stratus patches and thin stratus above. Tops of cumuli, 9500 ft. ; stratus patches, 8000 ft. ; thin stratus 

 above, 10,000 ft. Clouds joined to shower on right. ' Taken, from about 8500 f 



Brooklands, 2 p.m., August so, 191 7. 



owers usually above 

 NO, 2527, VOL. 



, and 

 to, 000 



lOl] 



the top of heavy 

 ft., though if the 



face, sometimes joined to the rain-clouds.' Some- 

 times also the lower clouds gather into large 

 cloud-heaps, which cause heavy showers in 

 the middle of the other rain-clouds. The 

 high rain-clouds may themselves be broken 

 iiilo shower'^ bv local vertical movements, so 



