ON THE MAGNETIC INTENSITY OF THE EARTH.

3

stances required to be known in order to estimate rightly thevalue of the results. In the case of observations which areeither wholly or partly new, these particulars are not to befound elsewhere; and in the case of those series, the publishedaccounts of which are contained in foreign works rarely metwith in this country, it has appeared desirable,—whilst givingevery direction which may facilitate a reference to the originalpublication,—to make the account here given complete in allparticulars essential to a just estimation of the value of theresults, independently of such reference. The details neces-sary for this purpose may render this portion of the reportoccasionally tedious to the general reader, who will be princi-pally interested by that section which contains the general con-clusions.

Section I.— Historical Notices.

It is to France we owe the first rightly directed experimentalinquiry on this subject. The instructions, drawn up by themembers of the French Academy of Sciences for the expeditionof La Perouse, contain a recommendation that the time of vibra-tion of a dipping needle should be observed at stations widelyremote, as a test of the equality or difference of the magneticintensity; suggesting also with a sagacity anticipating the result,that such observations should particularly be made at thoseparts of the earth where the dip was greatest and where itwas least.

The experiments, whatever their results may have been, whichin compliance with this recommendation were made in the ex-pedition of La Perouse, perished in its general catastrophe;but the instructions survived, and bore fruit in the earliest re-corded observations of the variations of the magnetic intensity,which are those published by M. de Rossel in the second volumeof the Voyage de Dentrecasteaux in search of La Perouse.

Rossel, 1791-1794.—These observations, though made in theyears above-mentioned, were not published until 1808. Theywere made with a needle vibrated in a dip circle of Le Noir,coming to rest disadvantageously soon for the purpose of experi-ments on the intensity. The needle continued in vibration littlemore than three minutes; consequently incidental errors wouldbear a very large proportion to the total time of vibration; adisadvantage which appears to have been in a great degree coun-teracted by the very great care bestowed on the observation.The needle was vibrated at Brest in 1791, before the voyagecommenced; and, successively, at Teneriffe; Van Diemen’s Land,