369 



them, or they were masked hy greater changes in the supports. 

 When the level of contact was substituted for the microscopes, or 

 when Mr. Sa.xton's reflecting pyrometer was employed, these changes 

 became very perceptible, and it was necessary to resort to direct ex- 

 periment upon the materials of the bars themselves to obtain even 

 approximate results, and then to correct a small residual quantity by 

 applying a covering more absorbent of heat to one bar than to the 

 other. If such changes have not been perceived hitherto, it has been 

 because adequate means were not used to delect them. 3. The 

 lever of contact and level, first used, I believe, in the adjustment of 

 standard measures by Bessel, was applied to indicate the lengths of 

 the bars. The levels were so delicate, that several divisions upon 

 them made up a quantity entirely insignificant in the measurement. 

 The doubt which I had was whether the sensibility of the apparatus 

 had not been carried too far; this was, however, entirely removed 

 upon finding the rapidity and certainty with which it could be used. 

 The contact between two adjacent measures was between a blunt 

 knife edge and a plane of agate. 4. The trussed support for the bars 

 adapted to bearing the apparatus at two points only, and the tin 

 covering or tube which surrounded the whole, were similar to those 

 used by Mr. Borden, but differed entirely in the adaptation of them; 

 the bars moved freely on the trussed frame upon rollers, and were not 

 attached to the covering tube in which the trussed frame itself was 

 merely supported. The tin covering was conical and was doubled. 

 •5. The tressles admitted of the various motions required in placing 

 the apparatus, and the length of the whole about twenty feet, (six 

 metres), gave a weight which permitted easy and rapid transfer by 

 four men, when covered with several thicknesses of imperfectly con- 

 ducting material to keep the fluctuations of temperature within mo- 

 derate limits. The contacts were usually made in much less time 

 than the setting of the forward tressles for the measure. The follow- 

 ing statistics of the measurement of a base line on Dauphin Island, at 

 the entrance to Mobile bay, with this apparatus, will suffice for the 

 present to show that we have obtained a useful auxiliary in a geodetic 

 survey, especially when the difficulties of triangulation render ad- 

 visable the measurement of frequent bases. The greatest length 

 measured in the course of a day, in the final measurement, was one 

 hundred and eighty-three tubes, equal to nearly seven-tenths of a 

 mile ; the least forty-seven, or one-quarter of a mile ; the average 

 one hundred and four tubes, or four-tenths of a mile. The whole 

 measurement of nearly seven miles was completed in seventeen 



