252 REPORT — 1861. 



Division I. — Dip. 



In selecting an instrument to be employed, in a magnetic survey, for the 

 purpose of determining the position, direction, and distance apart of the 

 isoclinal lines, care must be taiien, 1st, that its " probable error" be small, so 

 that the observations made at the different stations of the survey may be 

 as far as possible comparable with each other; and 2nd, that what is usually 

 termed the " Index Error " be either so small as to be practically insignificant, 

 or that, if of significant amount, its value should be carefully determined by 

 a sufficient examination at a base station, so that the general conclusions of 

 the survey may be comparable with those of similar surveys made in other 

 countries at the same epoch. The instrument selected for this survey was 

 one of the well-known English pattern which has been adopted for some 

 years past at the Kew Observatory; the circle was 6 inches in diameter, 

 fitted with both verniers and microscopes, and with two needles, each 3| 

 inches in length. An examiriation of the results obtained with twelve circles of 

 this pattern with their ^^ needles in 282 determinations made by different 

 observers at the Kew Observatory, has been published in the 11th volume of 

 the ' Proceedings of the Royal Society,' pp. 14'5-162. The circle there 

 distinguished as No. 30 was the one selected for the English survey, and 

 ■was employed at all the stations of observation in 185S, 18,59, and 1860. In 

 the autumn of I860, the English survey being tiien supposed to have been 

 completed, No. 30 was sent to the Magnetic Observatory at the Isle Jesus, 

 near Montreal in Canada, on the a])plication of Dr. Smallwood, Director of 

 that observatory ; but the pressure of other avocations having obliged me 

 to defer for a few months the preparation for the press of the results obtained 

 in 1858, 1859, and I860, I was enabled in the summer of 1861 to add four 

 more stations on the eastern coast of England, and for these observations 

 I obtained from the Kew Observatory the use of the Circle No. 3f?, which 

 (with its two needles) had been one of the twelve employed in the- com- 

 parison at Kew, of which the account is publi>hed in the ' Proceedings of 

 the Royal Society' as above stated. Referring to that account, it will be 

 seen that 28 of the 282 determinations at Kew with the twelve circles were 

 made with No. 30, and 4-9 with No. 33 ; and after the proper corrections for 

 secular change and annual variation, we find the " probable error" of a single 

 determmation to be with No. 30 ± r-25, and with No. 33 + r-18 ; whilst the 

 "probable error" derived from the 282 determinations obtained with the 

 twelve circles is +l''45. We may therefore regard Nos. 30 and 33 as instru- 

 ments superior rather than inferior, in the intercomparability of the results 

 obtained with them, to the average of their class; which class is, I believe, 

 unsurpassed by any other form of instruments in use either in our own or in 

 any other country for the determination of the magnetic dip. 



In regard to the question whether any and what " index correction " should 

 be applied to the results obtained with Nos. 30 and 33, it may be seen by 

 an examination of the results in the 'Proceedings of the Royal Society' 

 already referred to, that the mean result of the 28 determinations at Kew 

 with No.30 exceeded the mean ofthe282 determinations with the twelve circles 

 by +0''8, and the mean of the 49 determinations with No. 33 exceeded the 

 mean of the 282 by +0'"7. These differences have appeared too small to 

 require the assignment of a specific index correction to Nos. 30 or 33 ; it is 

 sufficient to record the circumstance, and to remark that it is possible that 

 the values of the isoclinal lines at the epoch of January Jst, 1860, given in 

 the present memoir, which rest on the results obtained witii these circles, 

 may be a fraction of a minute, or even a whole minute too high. 



