44-6 Indications of the Barometer and Thermometer 



hausted, a new polarization commences, to be added in its turn 

 to the initial current ; and the reason why we get the increased 

 work at the voltameter is, that 'while the polarization is pro- 

 ceeding at a c water is decomposed in the voltameter, and 

 although this may be somewhat less than the battery would 

 produce without the interposed plates a c, still this deficiencv 

 is more than made up by the action of the double pair at each 

 alternation of the wires. If the view I have taken be correct, 

 as reaction can never be greater than the action which occa- 

 sions it, we should never get, in Experiment 3, beyond the 

 quantity of gas given by two pairs of the battery d b, but we 

 may indefinitely approach that maximum. 



A commutator might be easily arranged instead of the hand 

 for effecting the alternation at the proper periods, which, by 

 a little contrivance, may be made to work by the battery 

 itself, but I prefer stating the experiment in its most simple 

 form and free from mechanical complication. 



Although the experiment as here described is merely in il- 

 lustration of a principle, it appears to me to promise results of 

 some practical value; the ceconomy of this method of apply- 

 ing force is evident, we get all but a double product with a 

 single consumption; the principle in all probability is not con- 

 fined to the voltaic force, but may perhaps be applied to 

 mechanics. 



LIV. Occasional 'Notes on Indications of the Barometer and 

 Thermometer during Stormy Weather at Belfast) from No- 

 vember 1833 to January 1843. By the Rev. William 

 Bruce*. 



1833. Nov. 28 /~)NE of the severest storms that had been 

 and 29. ^~* recollected. Wind began at about south- 

 east, changed during night to north-west or west-north-west 

 with tremendous rain. Barometer rose from 5 o'clock p.m. 

 of November 28, full one inch and a half in thirty-six hours, 

 having previously fallen with great rapidity, but the quantity 

 not noted. By the Library Barometer there was a rise of 

 1*37 in forty-five hours, viz. from 2 o'clock p.m. on November 

 28 to 11 o'clock a.m. on November 30, but no account could 

 there be taken of the additional fall from 2 o'clock to 5 o'clock, 

 when the rise began. 



1 834-. Nov. 5. — Barometer continued fallingduring the whole 



day till 1 1 o'clock p.m. On the morning of the 6th it had risen 



about four-tenths, and continued to rise till 11 o'clock p.m. (a 



beautiful day) ; on the morning of the 7th it had fallen nearly 



* Communicated by the Author. 



