246 Mr. Michael Donovan on certain Improvements in 



In this state of extreme sensibility the needle is subject to making unac- 

 countable excursions, amounting to 5^, 10', or 12' east or west. On one occa- 

 sion, when I had suspended a needle by a new silk fibre, it hung for that day 

 exactly north and south. Next morning, at five o'clock, it was found 8'; at nine 

 o'clock it was 0, and so remained all day. These changes recurred every day, 

 about the same liours, during the month of April. On some days, the excursion 

 and return took place twice. About the fifth week, the needle stood constantly 

 at 8° for two days ; it then shifted to 10° in the opposite direction, but returned 

 to in the middle of the day. I endeavoured to trace these changes to torsion 

 by an hygrometric quality of the silk, or to alterations of temperature, but 

 could not come to any certain conclusion, although I still attribute them to 

 either or both of these causes, knowing no other that could operate. 



Be this as it may, these variations in the direction of the needle would be a 

 source of false estimation of any deflective force, were the galvanometer planted 

 according to the indications of a needle thus in error: and variations will more 

 certainly occur, the greater the sensibility of the needle. The remedy, however, 

 is easy: the pin which sustains the silk fibre may be turned round ever until 

 the needle point accurately; but this mode of rectification is only admissible on 

 the condition that the position of the galvanometer has been rectified by its 

 independent compass needle, and its magnetic meridian line: and here is 

 another use of this needle and line ; without them, what errors might in such 

 cases be committed. 



I have frequently found, that when the compound needle was adjusted ex- 

 actly north and south by the volta-magnetometer, it pointed 10° or 12' diffe- 

 rently when carried to a different room where the galvanometer was stationed 

 to receive it: hence the rectification of the needle by the volta-magnetometer 

 should be effected beside the galvanometer. 



From what has been said, it is evident that the volta-magnetometer is only 

 necessary when needles of exceeding sensibility are required. In ordinary 

 cases, the method of magnetizing already described, or that commonly practised, 

 is sufficient. 



Havins now brought under notice the erroneous bearing which the com- 

 pound needle is apt to assume, and the possibility of its being in error, even to 

 the amount of 90°, when the deviation of one bar from the vertical plane of the 

 other is so small as to be undiscoverable by the eye, it is obvious how bad a 



