170 Transactions of the South African Philosophical Society. 



sions have been selected from the records of the six years 1898- 

 1903, namely, 



J. F. M. A. M. J. J. A. S. 0. N. D. 

 6, 6, 5, 7, 11, 11, 7, 13, 10, 10, 10, 9, 



and the hourly elements considered of the day upon which the centre, 

 or trough, of the depression passed over, together with those of the day 

 before and the day after." Depressions in which the centres belong as 

 much to one day as to another (i.e., when of any four days contain- 

 ing the centre the pressure on the second day is about equal to that 

 on the third, and the pressure on the first about equal to that on 

 the fourth), have not been taken into account. They are equally 

 important with the others, and will be considered at some future 

 opportunity. In selecting our 105 depressions no attempt has been 

 made to distinguish between one type and another, whether they be 

 primary or secondary cyclones, " Vs," or cols. Such a distinction, 

 which belongs more especially to the province of synoptic meteor- 

 ology, must also be left over for the present. 



It is evident from Table 1 that the average pressure on the first 

 days comes out, as it happens, almost exactly equal to the average 

 pressure on the third days. This is so far fortunate that it gives us 

 three quite central days in an average typical depression to study. 

 It appears, further, from the Table, that the average pressures on 

 the day before and the day after are about equal to the normal means 

 throughout the summer half of the year, but that the normals are 

 considerably the greater during the winter half.f The inference 

 seems to be that, on the whole, during the summer a depression 

 takes about three days to pass over, but that during the winter the 

 depressions are larger and take longer to pass over. For in the 

 winter months the front of a depression is evidently well defined 

 upon the day before the passage of the centre. 



In Table 2 the mean hourly values of pressure during the 

 passage of a depression are given, and also the deviations from the 

 normals hour by hour. It is remarkable that the depression does 

 not in the least obliterate the ordinary diurnal variation. The 

 absolutely lowest pressure comes about 4 p.m., and deviates at the 

 same time by the greatest amount from the normal. Since the 



* These may be conveniently distinguished as the first, the second, and the 

 third days. 



f The monthly normal means of the various elements used in the text are taken 

 from " An Elementary Synopsis, &c," in Trans. S. A. Phil. Soc, vol. xiv., 

 part 2. The hourly normals are from " The Determination of Mean Kesults, &c," 

 in Report of the S.A. A. A. S., vol. i. The first are obtained from four years' 

 observations, the second from five. In consequence the means of the monthly and 

 hourly averages differ a little. 



