22 Description of an Automatic Printing Brometer. 



are carried on, with but little trouble. In place of the 

 ivory pin usually employed, we would substitute a double 

 platinum pin ; one point of which should pass through a 

 small metal disk resting on the surface of the mercury in 

 the cistern, and the other, placed above the surface of the 

 disk. If these two pins are insulated, and connected with 

 the poles of a galvanic battery, when the mercury in the 

 cistern is raised, bringing the metal disk in contact with 

 the platinum point, the current will be established ; and 

 an audible signal can be given by means of an electro- 

 magnet. By this means we eliminate all errors and per- 

 sonal equation, in determining the zero of the cistern ; 

 and consequently the remaining error of reading will be 

 confined to the bisection of the upper surface. The actual 

 test of the magnetic connection as applied to our barometer 

 gave the following results. We would remark, however, 

 that no metal disk was employed, but one platinum wire was 

 plunged directly into the mercury, and the other was made 

 of the same length as the ivory pin, and only touched the 

 mercury when the , contact was made for bringing the 

 mercury in the cistern to zero. 



A large number of readings made by Mr. McClure and 

 myself, showed that by this method the personal equation 

 was entirely eliminated ; the mean of five or ten consecu- 

 tive readings not differing by an appreciable quantity. 

 The mean error for a single reading amounted to 0.001 of 

 an inch, and the maximum error to 0.002 of an inch ; show- 

 ing conclusively that the chief source of error in reading a 

 barometer lies in the adjustment of the surface of the 

 mercury in the cistern to the zero. 



