114 Transactions of the Royal Society of South Africa. 
possibly of some use to airmen. Yet it may be of interest to note that 
Clement Ley claimed ('Met. Mag.,' 1876) that the thunderstorms of Western 
Europe are divisible into three distinct classes. Abercromby claimed (B.A. 
^ Keport,' 1887) that there are at least three. Sir Napier Shaw remarks 
('Forecasting Weather,' 1911, p. 293) that "the conditions under which 
thunderstorms should be forecasted are as various as the types of thunder- 
storms themselves." Clement Ley put it in this way : 
" Those of the first class predominate in the winter months. They are 
far the most numerous on the western shores of Europe. They occur with 
low, or very low, barometrical pressure, and most commonly with steep 
gradients and strong winds, and are almost confined to the southern segments 
of depressions. They rarely take place when there is any very great diffe- 
rence between the direction of the upper and under currents. They are 
equally common by night and by day. The altitude of the storm clouds in 
this class is seldom so great as in the other two classes. Hail or snow 
frequently accompanies them. They are eminently local, and may, I think, 
be conveniently distinguished as thunder -squalls. 
" Storms of the second type are very rare, except in the warmer months. 
They are least common upon our western coasts. They predominate with 
light winds, with moderately low pressure, and in the southern segments of 
•depressions, but they are not at all particular as to the amount, or as to the 
distribution of pressure. They are local in character, but may be very severe. 
They occur almost exclusively by day. Their formation may often be readily 
watched ; cumulus, frequently under a sky devoid of upper cloud, swells ot 
,cumulonus ; the summit of the latter assumes the cirri-form aspect, and a 
massive local nimbus is developed. I would denominate them thunder- 
shoivers. 
" The third class are thunderstorms, par excellence. They belong to the 
warm season. They are remarkably rare on our extreme western coasts. 
. They are most common . . . when the isobars show local 
irregularities. But their most distinctive characteristic is the presence of 
an extensive and generally rapid southerly upper current (with more or less 
antagonistic surface winds) carrying cumulo-stratus over a wide area, 
especially to the northward and eastward of the storms. Storms of this 
class are as common by night as by day. They are usually less local than 
the others. They travel at a great altitude, are often exceedingly intense, 
and afford the grandest electrical display, but are productive of remarkably 
few accidents in proportion to the number of discharges. Their first develop- 
ment can rarely be watched to advantage ; but the process appears to be 
the reverse of that which occurs in the * thunder-showers,' the aggregation 
commencing in the higher regions of the atmosphere, and being propagated 
downwards." 
Comparing this account with my statistics, it would seem that Kimberley 
