M. F. Kohlrausch on a Variation-Barometer. 363 



Thus nearly 25 scale-divisions correspond to 1 inillim. of the 

 mercury barometer; so that the variations of pressure which 

 occur with us require a scale a metre in length. Sudden 

 opening or closing the door of the observing-room causes the 

 image to move over 2 or 3 divisions. A further augmentation 

 of the sensitiveness could be obtained without difficulty by in- 

 creasing the distance of the scale or placing the point of contact 

 higher, but might not answer any purpose. 



The instrument is, of course, intended for observations within 

 a brief space of time; nevertheless, in the course of an entire 

 month, mine changed its position relative to a mercury baro- 

 meter by only about \ millim. of the latter. 



The influence of temperature was uncommonly great, viz. 0*3 

 millim. for 1°. As a fall of the position corresponds to a rise 

 of temperature, it is evident that the phenomenon arises from 

 too much air being contained in the tube, and that it can be 

 easily obviated. Since we have, as two opposite corrections, the 

 diminution of the elasticity of the metal and the increase of that 

 of the internal air, it must be possible even to reduce to zero the 

 temperature-correction of any aneroid barometer by fixing its 

 content of air at a determinate quantity. From the series of 

 observations which I have made occasionally in the course of the 

 last few months it follows that, as was to be expected, the pres- 

 sure of the air seldom remains constant, even for a short time. 

 For the most part, very small variations are constantly taking 

 place. It is different, however, at times of high wind. The 

 upper curve, fig. 4, gives for instance the course during three 

 minutes on the 26th of February last, a stormy day. The 

 entire height of the figure corresponds to 1 millim. of mercury; 

 so that the greatest fluctuation denotes^ millim. In the ob» 

 serving-room a window, turned away from the direction of the 

 wind, was opened. The noting-down took place every three 

 seconds ; nevertheless a great number of small variations could 

 not be recorded. The lower curve exhibits the progress during 

 a moderate hail-shower accompanied by a thunder-storm on 

 May 30, likewise in the course of three minutes. The tempest 

 had commenced some minutes previously. The smaller undula- 

 tions which follow the initial larger wave continue, diminishing, 

 for some time. 



The past summer afforded frequent opportunities of observa- 

 tion during violent storms ; yet I have never been able to detect 

 any connexion between the oscillations and lightning. 



I must remark in conclusion, that M. Eontgen has employed 

 the mirror- reading in another way for measuring the pressure, 

 and also projected an aneroid barometer with mirror and scale*. 

 * Pogg, Ann. vol. cxlviii. p. 624. 



