66 Mr. G. J. Symons. On Barometric Oscillations [May 8, 



The fourth pen is actuated somewhat like a piano. On the occur- 

 rence of a flash of lightning, the observer presses a key, the pen 

 travels slightly to the right, and flies back to zero. Referred to the 

 automatic time-scale, this gives, to a second, the time at which the 

 key was depressed. 



The fifth pen is similar, but, being intended to record the thunder, 

 the observer will continue to hold down the key until the roll is in- 

 audible. The time of the departure of this pen from zero will 

 evidently be later than that for the lightning by the time-interval 

 due to the distance of the flash, and possibly something may be 

 learned from the accurate record of the duration of the thunder. 



The sixth pen is similar to the third, and is intended to record the 

 time, duration, and intensity of hail. 



The seventh and last pen is devoted to an automatic record of 

 atmospheric pressure. As the rapid motion of the paper, which is 

 indispensable for studying the details of a thunderstorm, has enlarged 

 the time-scale more than a hundredfold, it was imperative that the 

 barometric scale should itself be greatly enlarged. But the range of 

 the barometer in London is more than 2^ in., and no enlargement less 

 than ten times the natural (mercurial) scale would be of any use ; 

 hence a breadth of 25 in. of paper would be necessary, unless some 

 mode of shifting the indication could be devised. 



Several plans were tried, but finally a modification of Richard's 

 statoscope has been adopted, which is so sensitive that it will indicate 

 the opening or shutting of a door in any part of the house, gives a 

 scale of 30 in. for each mercurial inch (i.e., about three times that of 

 a glycerine barometer), and yet only requires 4 in. breadth of the 

 brontometer paper. Without entering into all the details of con- 

 struction, it is desirable to explain the general principle, and its 

 application. As it was essential that the apparatus should record 

 accurately to O'OOl in. of mercurial barometric pressure, it was 

 evident that friction had to be reduced to a minimum, and consider- 

 able motive power provided. This is done by placing in the base of 

 the brontometer a galvanized iron chamber, which contains about 

 3 1 cubic feet of air ; on the upper part are a series of elastic chambers, 

 similar to the vacuum boxes of aneroid barometers, but much larger. 

 When the instrument is to be put in action, these chambers are 

 connected with the large air-chamber, and a tap is closed which shuts 

 off communication with the external air. Any subsequent increase, 

 or decrease, of atmospheric pressure will compress or allow to dilate 

 the air in these chambers, and the motion of the elastic ones produces 

 that of the recording pen. 



Obviously, any large change in the temperature of the confined 

 air would vitiate the readings ; but (1) the instrument is not required 

 to give absolute, but merely differential, values, and (2) the influence 



