JAN. — MAR. 1857.] for the Provinces of Madras. 
thought and that the new gauge will not answer throughout every 
monsoon, as 8-25 inches fell on the 9th October 1852, and 10-7 
inches on the 8th October 1855. He thinks however that a few 
gauges of the new pattern would be useful and asks for 4. 
In his second letter No. 202 dated 8th July, he suggests that 
Mr. Bayley's improvement of dividing his measuring rod into 
lOths should be introduced with the present gauges and points out 
the method of using it. 
* 
The total number of the new gauges asked for was 129, but as 
fet>me of the Collectors who approved of 
, the new gauge did not mention the num- 
bers they required, I have the honor to 
recommend the manufacture of 180 
gauges of the new pattern, substituting 
however a tin vessel to receive the rain 
fall instead of the quart bottle ; or else . 
placing the bottle in a tin cylinder so that 
if very heavy rain fall, so as to more than 
fill the bottle, the surplus water might 
overflow into the tin vessel as exhibited 
in the annexed drawing. This would an- 
swer very well, as the external diameter 
of a quart bottle is less than that of the gauge ; the former being 
about 3i and the latter 4-697 inches. 
The total cost including that for the tin vessels in which tlie 
bottles are to be placed would amount to Rupees 360 for 180 
gauges. 
It is far from being an uncommon occurrence to receive mis- 
taken values of rain falls from Collectors using the pattern now in 
use, and I have frequently to send back the returns for correction 
month after month. The new stick suggested by Mr. Elton allud- 
ed to in his letter of the 8th July, and which I beg leave to forward 
for inspection is however a decided improvement, and to any one 
giving himself the trouble to think on the subject at all, it is diffi- 
cult to believe that wrong values can be assigned to the measure- 
ments shown by the decimal scale. But it must be admitted that 
