62 Transactions of the South African Philosophical Society. 



mean pressure of the month, thirteen when the pressure was 

 distinctly above the mean, the other fourteen at about the mean. 

 Many of the thirteen which took place when the barometer was 

 relatively high were still under a depression. If we divide the 

 mud rushes into two sets, large and small, we get the following 

 result : — 



Table 3. — Distribution of Mud Bushes on the Barometric 



Gradient. 



Large Small Total 



Rushes. Rushes. Rushes. 



On a downward slope 6 8 14 



In the pit of a depression ... 8 4 12 



On an upward slope 8 1 9 



Under a crest 4 4 8 



Under a steady pressure 3 4 7 



Uncertain 1 1 2 



Total 30 22 52 



At the same time some of those which are classed as under a 

 steady pressure may yet have been subjected to great and sudden 

 barometric changes, unobserved because of the great intervals 

 separating the consecutive observations. This is a matter which 

 could only be settled by a continuous barometric record, and, 

 unfortunately, the photo-barograph of the station under my charge 

 has not been running long enough to do so in every case. Certainly 

 on April 30, 1896, although the barometer stood nearly at its 

 normal levels at the standard observing hours, yet in between those 

 times the changes were very great. On one other day, by a piece of 

 bad luck, it happened that the photograph was spoiled by a tem- 

 porary failure of the clock. I should not feel inclined, though, to 

 put much faith in the statistics of Table 3, because much would 

 depend on one's notion of the weight to be given to the descriptive 

 terms, wherefore others adopting the same method might get some- 

 what different values in each class. It is safer to form a judgment 

 upon the actual curves. 



What seems to be a most significant fact is revealed by an 

 analysis of the hourly distribution of these rushes : Fifty-one per 

 cent, have taken place during the second shift, i.e., between 3 p.m. 

 and 11 p.m., twenty-seven per cent, between 7 a.m. and 3 p.m., 

 and the remaining twenty- two per cent, between 11 p.m. and 7 a.m. 

 In other words, the greater proportion come in the shift during which 

 the barometer stands at the lowest point of its diurnal range. The 



