86 On the Trigonometrical Survey 



as good an accordance as resulted from the two last measure- 

 ments of the Base at Ralph's Bay. 



I must mention one other circumstance connected with 

 these operations, and which at first led me to doubt the 

 safety of dependence upon the deal rods. 



The measurement of their length by the four- foot stand- 

 ard during the operations, especially at Norfolk Plains, indi- 

 cated a small amount of contraction and expansion not to be 

 expected, and in no way to be accounted for; but it appeared 

 that these measurements were made in the extreme heat of 

 mid- day, and arose from the steel standard being slow in 

 following the changes of temperature indicated by a detached 

 thermometer. The temperature of the metal, in fact, was not 

 ascertained at a high temperature of the atmosphere; and as 

 the rods embedded in their coflFers, and screened from the sun, 

 could undergo no such sudden changes, I rejected the mid-day 

 measurement of their length, and adopted those taken early, 

 when no great allowance was required to be made for the 

 effect of temperature on the standard. The Base at Ralph's 

 Bay being measured in cooler weather, this difficulty did not 

 arise. 



For the angular observations from the main triangles, an 

 i^ltitude and Azimuth instrument had been obtained from 

 England, with a repeating table of excellent finish, both 

 Tery portable, and at the same time efficient. 



The horizontal arc of the instrument is twelve inches in 

 diameter, graduated to 10." 



Many repetitions were made of every angle with reverse 

 observations, and every possible attention to ensure an accu- 

 rate mean ; and the result has been most satisfactory. 



The greatest error in the sum of the angles of one tri- 

 angle was 3"3 seconds, and this was in a triangle of nearly 

 forty miles sides. 



