214 GEODETIC OBSERVATIONS. 



but owing to a defect in No. 2 transit telescope similar to that found 

 in 1872-73, three of these arcs were remeasured in the following year 

 as well as five new arcs, and in 1883-84 the operations moved east- 

 wards of Calcutta as far as Moulmein, thus completing the last 

 link between Karachi and Moulmein. 



In 1884-85 Major Heaviside resumed the observations for 

 astronomical latitudes with the zenith sector, which had been in 

 abeyance since 1871-72. He observed at stations about half a degree 

 apart down the Amua meridional series, commencing about latitude 

 28° 30' N. and carrying them south over five degrees of the arc, 

 which was eventually to be extended to Madras in latitude 13° 4'. 

 The results are given in the appendix to the Survey Eeport for 

 1885-86, p. lviii. 



No fewer than nine arcs were measured for longitude by electro- 

 telegraph in 1885-86, but, unfortunately, when the results came to 

 be reduced and compared in the recess season, it was found there 

 was a mysterious source of appreciable error, for out of five verifi- 

 catory circuits! three exhibited errors between a quarter and a third 

 of a second of time. These irregularities were investigated with 

 the greatest care during the following season by Colonel Haig, 

 Lieiitenanl S. CI. Burrard, R.E., and Mr. Eccles, and every possible 

 interchange of observers, telescopes, electrical apparatus, and stations 

 was secured in a series of observations on a short experimental arc 

 at Dehra Dun. The observations tended to show that the longitude 

 operations had been harassed by three sources of error, viz., (1) 

 electrical, (2) local, and (.'!) instrumental. But it must be borne in 

 mind that these so-called errors are intrinsically so minute as to be 

 of no consideration whatever, except from a purely scientific point 

 of view.J 



* The whole of the details ami results of the observations from 1875-76 to 1883-84 

 will be found in Vols. IX. ami X. of the "Account of the Great Trigonometrical 

 •• Survey," &c. mentioned in tlie previous note. 



t Any three ares forming a triangle otter a verificatory circuit: 

 thus the sum of the ares A B and 15 C is equsl to the arc A C, 

 therefore, if the sum of the measured arcs A B and 1! C differs 

 from the measured are A C the difference is due 1o error of 

 measurement in one or more arcs. 



;'; They have since been traced to the determinations of the errors ofcollimation of the 

 telescopes, which indicate either slight weakness in the telescope tubes, or irregularities 

 in the object glasses, which were not always exactly centered between the collimators 

 in determining the collimation. 



