1838.] Description of Sketches of a Self-Registering Barometer. 1'25 



tached, and communicates with the vernier in front through the elongat- 

 ed narrow opening, as shown in the sketch. F is the scale of inches, 

 which may consist of two or more parts, — the inches being numbered 

 in succession ; for if of only one piece, it will be longer than the space 

 between GG, and of course project considerably to the left, H is a 

 joint to admit of the part above it being inclined to the position most 

 favourable for the reading off. 



In a former communication it is stated, that when the register baro- 

 meter is put in operation, the pencil is set at the line in the middle of 

 the roller (for the sake of distinction called zero line) ; and that the 

 height of a common barometer at the time is to be noted. Suppose 

 this height to be 28 inches at 10 a. m., when it is intended to read off 

 the pencil trace, it is evident that the vernier and scale ought to indi- 

 cate 28 inches, when the former is brought to touch the 10 a. m. hour 

 line. This adjustment of scale and vernier is accordingly to be made, 

 when the latter is to be secured by the screws in GG under which it 

 passes. 



The register roller, as formerly explained, registers only the varia- 

 tions of atmospheric pressure, that occur above or below the average 

 barometrical height of the place where it is set in operation. The 

 longer the expanded top of the tube is, the instrument will of course 

 embrace a proportionally greater number of stations of various rela- 

 tive height, without the necessity of moving the cistern 'D,fig. 1. But 

 greater length of expanded top of tube, than what is required for baro- 

 metrical range, is quite unnecessary, as the adjustment of the surface 

 of mercury to the middle of expanded portion of tube can easily be 

 effected by raising or lowering the cistern D fig. 1 or the cistern E 

 fig. 2. Thus at Madras the barometer ranges about 30 inches — on the 

 Neilgherries about 23. The cisterns therefore must be raised about 

 seven inches to admit of their respective instruments being used at the 

 latter station. 



