6 



Rev. S. J. Perry on Magnetic Observations [Jan. 10, 



are necessarily affected by many possible sources of error, when the 

 work has to be done in the course of a long and difficult journey. The 

 carriage of the instruments by land and sea ; the frequent packing and 

 unpacking ; the great variations of temperature ; the shortness of 

 time at each station, scarcely allowing sufficient attention to be paid 

 to choice of locality or examination of instruments ; all tell more 

 against the magnetic declination than against the other elements of 

 terrestrial magnetism. 



As we were detained a considerable time at the Cape of Good 

 Hope, it was thought advisable to examine thoroughly all the instru- 

 ments, and to take a very complete set of observations at the Cape 

 Observatory, in order to make this well-known station the point of 

 departure for all our work in Kerguelen and elsewhere. It was 

 fortunate that this could be done leisurely, as we found on examina- 

 tion that the declination magnet was useless, the bar not being 

 magnetic, nor capable of retaining a sufficient amount of magnetism to 

 secure accurate results. On July 30th the bar was magnetised by aid 

 of large permanent magnets, and then subsequent declination observa- 

 tions led to most unsatisfactory results. On August the 19th a coil was 

 made use of to magnetise the same bar, but two series of readings then 

 taken with the needle, and with the Observatory magnet, shewed the 

 utter untrustvvorthiness of our needle. As the Rev. W. Sidgreaves, 

 though most kindly assisted by the Government Astronomer, Mr. E. J. 

 Stone, was unable to magnetise satisfactorily our reversible declination 

 needle, we were forced to substitute in its place the vibration magnet. 

 The latter not being arranged for rapid reversal, the zero of the scale 

 was very carefully determined both at the Cape Observatory and after- 

 wards at Kerguelen, and the scale value accurately tested. 



The method of observation usually adopted, was to fix the position 

 of a well defined distant mark, by aid of sun or .star transits, with the 

 Simms' theodolite, and then to note the bearing of the magnetic needle 

 with respect to the same mark with the Jones' unifilar. At Florence 

 the confined space, and still more the badness of the weather, made it 

 advisable to trust wholly to the unifilar, and therefore no fixed mark 

 was used, but the position of the magnet was referred directly to that 

 of the sun, observed by aid of the mirror attached to the instrument. 

 The same method was used at Moncalieri, and the corrections for the 

 displacement of the mirror, deduced from the latter observations, 

 served to correct the single reading at Florence. The necessary 

 correction for any error in the mirror is explained in a former paper 

 on the magnetic survey of the west of France (Phil. Trans. 186*9, 

 p. 43). 



The chronometer made use of throughout the whole series of ob- 

 servations was a good pocket instrument from Greenwich Observatory, 

 Maurice 6144. It was compared on every possible occasion with the 



