14 Mr Harris on Magnetic Intensity by the 



cuo in small arcs, 100 vibrations in 11' 56", 6, after exposure for 

 about half an hour to a cold of about 5° of Fahrenheit's scale ; 

 it made under similar circumstances 100 vibrations in 11' 65" ; 

 whilst in other cases, similar bars, after being freely exposed to a 

 range of temperature from 0° to 212°, appeared to remain inva- 

 riable, in respect of certain ranges of temperature taken between 

 these points. These results have received some confirmation 

 since the notice of my paper on this subject, read at the Meeting 

 of the British Association in June last, as appears in a letter to 

 Dr Brewster from Professor Kupffer of the Imperial Academy 

 of St Petersburgh, printed in the London and Edinburgh Jour- 

 nal of Science for August 1832. The coincidence between Mr 

 Kufffer's results and my own is not a little remarkable. I 

 have been hence led to hope, that the effects of heat in disturb- 

 ing the tension of a vibrating needle, may in certain cases be con- 

 fined to differences of temperature, without the limits of that to 

 which the needle has been previously exposed. I desire, however, 

 to evince no greater confidence in this opinion, than the results 

 of the experiments above alluded to seem to warrant. Mr 

 Christie's profound and interesting researches in this depart- 

 ment of science have undoubted claims to our confidence ; and 

 we must therefore endeavour, by further investigation, to discover 

 the source of the apparent anomalies which occasionally present 

 themselves. It may be fairly observed, at least, that whilst such 

 questions as these, and others of no less consequence, remain in 

 any degree doubtful, it is quite impossible to place confidence in 

 results pretending to detect small periodical variations in the ter- 

 restrial intensity, which variations, if such exist, may cause diffe- 

 rences in time, certainly much smaller than might alone arise from 

 either of the foregoing causes. 



25. Beside the changes of tension induced in a bar by heat, 

 we have to consider those which may happen from the manner 

 of preserving the magnets themselves, such as occasional contact 

 with each other, or with masses of iron. The usual method of 



