30 SIXTH REPORT.— 1836. 



The author of this communication also mentioned some experi- 

 ments illustrative of the general advantage of temper in bars or 

 needles in a moderate degree of hardness throughout their length, 

 instead of being tempered, as they usually are, both in sea compasses 

 aiid in ordinary magnets, only at the ends. This result, which at 

 first sight might seem at variance with those obtained by Captain 

 Kater in his laborious investigations for determining the best con- 

 struction for sea compasses, the author showed was not inconsistent 

 Avith established principles ; for whilst he admitted the correctness 

 of Captain Kater's conclusion as to the superiority, in point of ori- 

 ginal energy, in needles tempered only at the ends, he suggested that 

 the ultimate advantage in long voyages, where great permanency 

 is requisite, or under circumstances where the permanency of the 

 energy is much tried, would probably be found in favour of tem- 

 pered needles. At all events, in regard to compound needles and 

 compound magnets, the author had abundant experimental evidence 

 to prove that a thorough tempering is absolutely necessary for the 

 adequate retention of the advantage gained by the combination of 

 bars over single bars of equivalent mass. 



Experiments on Terrestrial Magnetic Intensity, especially in relation 

 to the Influence of Height. By Professor Forbes. 



These experiments were made with Hansteen's apparatus, the 

 property of the Royal Society of Edinburgh, chiefly in the years 

 1832 and 1835. The author particularly proposed to himself the 

 problem of the influence of height upon intensity, considering the 

 observations of KupfFer to be quite inconclusive as well as those of 

 preceding experimenters. He showed that by choosing stations at 

 considerable elevations, and placed on a ridge so as to have compa- 

 ratively low stations on either hand, the influence of geographical 

 position in affecting the results may be eliminated ; the intensity at 

 the lower level for a point vertically below the elevated station being 

 obtained by interpolation, the diff'erence between it and the observed 

 intensity may be fairly attributed to the influence of height abstract- 

 ing from local disturbing causes. To correct for these, and likewise 

 to attain considerable numericeJ exactness, multiplicity of observa- 

 tions is most desirable, nor can any satisfactory result be looked for 

 from a single experiment. It appeared from the tables of Professor 

 Forbes's observations in the Alps and Pyrenees, that the sum of the 

 heights to which the Hansteen apparatus has been carried by him, 

 and which forms the basis of his induction, is more than 1 60,000 

 feet, or 30 vertical miles. 



The author stated that he had not yet submitted his observations 

 to one system of calculation from which to deduce the elements of 

 disturbance with the greatest accuracy ; but he pointed out from a 

 great number of individual observations that had the diminution due 



