572 VHILOSOPHICAL TKANSACTIOMS. [anNTO I768. 



the breadth of the bottom board was 7-^ inches, that of the top = 3 inches, of 

 the ends = 44- inches, and the bottom and the top were strengthened with 

 boards firmly fixed to them at right angles. The joints were secured with plates 

 of iron, and the ends were plated with brass. The plumb-lines used in setting 

 them level, were = 3 feet and 2 inches in length, and hung in the middle of the 

 levels, being secured in a tube from the wind, in the manner of carpenters' levels; 

 on which account we called these by the same name. When the plumb-line 

 bisected a point at the bottom, the ends were perpendicular. Where the ground 

 was not horizontal, or there were logs, &c. to pass over, one end of the level 

 was raised by a winch and pully. 



The level being set, a short staff was driven into the ground, very near and 

 opposite the plumb-line, in the top of which moved a thin plate of iron, about 

 12 inches long; at the ends of which were points, directed to the intersections 

 of lines, drawn on the board that covered the plumb-line. By bringing the 

 points in a line with one 'of the said intersections, if the level was by accident 

 moved, it might be discovered, and brought again to its place. 



A level being thus marked, the end of the other was brought in contact with 

 it, and marked in the same manner, before the first was moved ; the first was 

 then taken up, and set before the last. And so the operation was continued. 

 Mr. Dixon attended one plumb line and staff, and Mr. Mason the other. 

 The measure was carried on in a straight line, and in the proper direction, 

 by pointing the levels to the farthest part of the visto that could be seen; this 

 was readily and accurately done, on account of their lengths. The levels were 

 frequently compared with the brass standard, of 5 feet, provided for that purpose, 

 and the diflference was noted between 8 times the brass standard, and the length 

 of the two levels taken together. This difference serves for reducing the 

 measure taken with the levels, to what it would have been if it had been taken 

 with the brass standard itself. For facilitating this comparison of the levels 

 with the brass standard, pieces of brass were fixed into the bottom boards of the 

 levels on each of which was drawn a faint line. And -^ of an inch at the end 

 of the brass standard being divided into 10 parts or lOOths of an inch, the 

 difference between 8 times the brass standard, and the two levels joined together, 

 was witli the help of a magnifying glass of a short focus, determined to great 

 accuracy. The brass standard being liable to alter with the changes of heat and 

 cold, a further correction becomes necessary on that account, in order to reduce 

 the measures to the temperature of 62° of Fahrenheit's thermometer, which is the 

 term to which the former operations of this kind have been reduced. For this 

 purpose, the rate of expansion of brass is taken from Mr. Smeaton's experiments, 

 made with a pyrometer of his invention (see Philos. Trans., vol. xlviii, p. 482, 

 vol. X, of these Abridgments) which is -rTnrVV of an inch on a length of one 



