176 OBSERVATIONS OF THE TERRESTRIAL 



The mean time of 200 vibrations, deduced from these results, 

 is 1137"'66 at the temperature 60'6. But the time observed 

 on the 18th of June is 1135"'87 at 61-0 ; so that on this day 

 the rate of the needle was less than the mean by 1 s - 79, a difference 

 which corresponds to an increase of '003 in the horizontal force. 

 The observation of the 5th of July exhibits a difference somewhat 

 greater on the other side. 



6. The last source of error which requires to be noticed under 

 this head is the change of the magnetic condition of the needles 

 employed. Independently of the derangements of magnetic 

 equilibrium induced by the presence of iron, or other disturbing 

 causes, it is well known that most needles lose something of their 

 original force. This loss is greatest at first; and the needle, if 

 originally well tempered and then magnetized to saturation, is 

 usually found to arrive at a nearly settled state in about a year. 

 Most of the cylinders employed in the following observations 

 seem to have reached that condition ; and the changes of mag- 

 netic state which they have exhibited are, except in the case of 

 Cyl. S (b), unimportant. In order to detect any such changes, 

 and to correct for them if they arise, it is necessary to observe at 

 the place chosen as the base of reference, at the termination of 

 each series of observations, as well as at their commencement. 

 If it is then found that the needle has lost any small portion of 

 its force, or if the time of vibration has augmented, the amount 

 of the correction due to each result may be found by assuming 

 the change to have been regular, or proportional to the time elapsed. 

 When the loss is very small, however (as was the case in the obser- 

 vations which form the subject of this paper), the correction may 

 be disregarded, provided we take as the time of vibration at the 

 base of reference the mean of the times observed at the commence- 

 ment and end of the series. 



The needles used in the present series are the cylinders L (a], 

 L (ft), made by Dollond, and belonging to Mr. Lloyd ; cylinder 

 S (b) belonging to Captain Sabine, and cylinders E (e) and B, (d) 

 in the possession of Captain James Boss. All the circumstances 

 of the observations are given in the annexed Table. The first, 

 second, and third columns contain time place, day of the month, and 

 hour of the observation. In the fourth column is set down the 

 observed time of 100 vibrations, or the immediate result of obser- 

 vation divided by 3. The fifth column contains the terminal an; 



