Meteorological Committee for 1899. 11 
40 years’ average of each of the same 10 years, with the result of 
a deficiency of only 5} in. This result is startling; but if the 
table be examined it will be seen to be the result of 3 years of 
excessive rainfall (for the excess in 1897 may be disregarded) ~ 
combined with smaller deficits in other years. I do not suppose 
that anyone would have imagined that over the series of years 
shown in the table the deficit at Surbiton would be nearly four 
times as great as at Wimbledon. 
Let us now come back to the year 1899, and see how the deficit 
shown in the first set of tables has arisen. 
With this view I have again constructed, on the model of 
previous years, tables A, B, C,and D. Tables A and B refer 
to Greenwich. It will be noticed that in table A there are only 
four months showing an excess of rainfall, whilst in table B 
there are five, the fifth month being September, the other four 
in each case being January, February, April, and November. 
Table C refers to Surbiton, and table D to Wimbledon. In the 
two latter tables the same five months as in table B are in 
excess. If one examines these tables some very curious features 
present themselves. I will take, first, the month of April. In 
table A, with the 80 years’ mean, there is an excess of 1:35 in. ; 
and in table B, with the 40 years’ mean, of 1°39 in., an excess 
which, though the same 40 years are taken, diminishes, in the 
case of Surbitor to 0°80 in., and Wimbledon to 0:73 in. Let us 
now take the month of November. This month was an ex- 
ceedingly curious one, being extremely wet at the beginning and 
extremely dry at the close. I have no doubt that you all, like 
myself, were wondering what the rainfall of that month was 
going to amount to, when it suddenly stopped completely. If 
we turn to the tables, we see that there is in table A an excess 
of 1-36 in., in table B of 1°54 in., in table C of 2:05 in., and in 
table D of 2:17 in., very nearly the exact opposite of what occurred 
in April. If we now turn to the deficiencies, we shall first 
notice the month of June. Here the deficiency increases through- 
out the four tables: in table A it is 0-75 in., in table B it is 
0°84 in., in table C it is 0°93 in., and in table D it is 1:07 in. 
Again, let us take the month of August. Here we have a distinct 
change, in that the deficiency in the four tables is practically the 
same, only differing from one another by a very few hundredths ; 
in table A it is 1-98 in., in table B it is 1:95 in., in table C it 
is 1-97 in., and in table D it is 1:92 in. Again, if we take the 
month of July we have another change. In table A the deficiency 
is 1-31 in., in table B itis 1:13 in. In table C it is nearly an 
inch more than in table B, viz. 2:03 in., whilst in table D it is . 
less than that of table C, viz. 1:82 in. It is extremely difficult 
to account for these different excesses and deficiencies without 
intervening gauges, which do not exist, and which even if they 
