1845.] DISTORTION. OF THE ICE MEASURED. Ill 



and gutters were made so as to drain as far as possible the sur- 

 face water from the part under experiment. The theodolite 

 being stationed as usual over Q, and the vertical wire of the tele- 

 scope describing a great circle passing through the line QQ1 

 transverse to the glacier, an assistant (Balmat), directed by my 

 signals, bored a series of holes from two feet to two feet, forty- 

 five in number, with a common carpenter's centre-bit, and as 

 nearly as possible in the visual line. The holes, which were 

 ffrth of an inch in diameter, and about five inches deep, were 

 immediately occupied by wooden pins prepared for the purpose. 

 These pins were placed as nearly as possible in the visual straight 

 line, but from the nature of the operation, some errors were in- 

 evitable. The amount of these errors of position or zero of the 

 marks was immediately determined by causing the vertical wire 

 again to traverse the series, the assistant placing over the centre 

 of the head of each pin in succession the zero point of a scale 

 of inches divided both ways, and held parallel to the length of 

 the glacier, so that (the divisions to tenths of an inch being very 

 plainly marked, and divisible by estimation by the telescope) 

 the fundamental position of each pin was determined, and con- 

 sidered as + if in advance of the traverse line (in the direction 

 of the glacier's motion), and if behind it (or nearer the origin 

 of the glacier). The mere error of reading did not in any case 

 exceed ^V tn f an i ncn > though the uncertainty of centering of 

 the theodolite over Q might amount to j^h of an inch, or even 

 more. The two marks nearest Q had their positions deter- 

 mined by a thread stretched from the station-pointer of the 

 theodolite to the third mark, their distance being too small to 

 be distinctly seen by the telescope. 



The very same process, as regards the placing the zero of 

 the scale on the head of the pin and reading off, was repeated 

 on subsequent days, and the new readings, minus the funda- 

 mental readings, gave the apparent relative motion in the inter- 

 val. This apparent motion had to be corrected, exactly as be- 

 fore explained, for the rotation of the visual line due to the 

 translation of the fundamental point Q. [Some details are here 

 omitted.] . . 



