CLOCK, CHRONOMETERS, AND TRANSIT. 453 



At Supply Bay a pier of brick and cement was built, on a foundation of 

 concrete, on hard ground; but Lieut. Corbet placed between the masonry 

 and the iron stand a block of wood, to which the iron stand was screwed ; 

 apparently to facilitate fastening down the latter. 



On November 20 a furious gale blew off the roof of the observatory, and the 

 transit was knocked out of its Y's, and fell to the floor on its eye-end and 

 setting circle, which was much bent. This being straightened with the 

 hands, the instrument was found to be apparently but little the worse for 

 the accident. 



The correction for inequality of pivots was carefully determined. It was 

 found to be 0"'ll (+ lamp B., lamp W.), which has been applied to all 

 the level errors given in the following tables. 



The error of collimation is determined by observations of stars with 

 reversed positions of the transit axis. It appears that Lieut. Corbet was in 

 the habit of testing the stability of the collimation and azimuth errors, by 

 directing the telescope to and reversing it on a distant terrestrial object ; but 

 the instrument having no micrometer, he could not of course make accurate 

 observations in this manner. He appears to have had no suspicion that 

 anything was wrong with the instrument, but was puzzled at the apparent 

 instability of the azimuthal position. 



I have no doubt that in this case, as with all the other transits made by 

 Troughton and Simms, the ring clamp at the eye end of the telescope, which 

 fixed the wire plate, had worked loose during the voyage, and consequently 

 the system of wires was not rigidly fixed with reference to the optical axis. 

 It is of course possible that the object-glass was loosely attached, but it is 

 improbable, from the mode of attachment. No record is to be found of any 

 special examination of the instrument. 



With the instrument in this condition, with the additional fault of being 

 mounted on a block of wood, in a climate where heavy rains and fierce winds 

 perpetually alternated, good observations for local time were impossible. 



The value of the divisions of the hanging spirit-level was determined by 

 the makers before leaving England, and again on the return of the expedition 

 to Greenwich ; 40 divisions were found equal to one minute of arc. The 

 level error was generally large, but the determination of local time is not 

 injuriously affected on this account. 



It follows from the description of the instrument and its mounting that 

 the azimuth error can never be inferred with any certainty. It has been 

 computed in every possible way from the observations. 



