BASE LINE MEASUREMENT. 43 



relative lengths at a normal temperature is, theoretically, an indication of the 

 temperature of the bars at any time. The arrangement for indicating their 

 relative lengths forms a part of the apparatus, and is intended to indicate 

 the temperature of the bars, and thus t< > affi >rd means of reducing the lengths 

 of the bars to a normal temperature. It has not been found, however, to 

 work well in practice. Besides this, there are other objections to the use of 

 bars of any kind, which may be summarized as follows : First, then use is 

 expensive. A considerable number of men are needed, and as the measure- 

 ment proceeds slowly it often requires from a month to six weeks to measure 

 and remeasure a base five miles in length. Again, since these bars are but 

 four meters in length, there are many contacts to be made in. each mile of 

 measurement, and each contact affords the possibility of a trifling error. 



In view of these objections and of certain positive advantages which 

 the change would produce, it was decided, in 1887, to drop the use of bars 

 in the measurement of base lines, and to adopt in their place long steel 

 tapes. By their use it has been found easy to attain the required degree of 

 accuracy in measurement, inasmuch as the number of contacts is reduced 

 to a small fraction of the number necessary in the use of bars, while the 

 uncertainty in regard to the temperature of the measuring apparatus is 

 reduced to a minimum by carrying on the measurement at night or in cloudy 

 weather. The expense of the measurement is greatly reduced. Fewer 

 men are required. The work of preparing the ground and the work of 

 measuring are much lessened, and the rapidity of measuring is increased 

 manyfold. The diminished cost makes it practicable to measure much 

 Linger bases, thus, diminishing the number of stations required in the 

 expansion. It allows, also, a measurement of base lines at shorter intervals 

 in the triangulation. 



The tape in use has a length of 300 feet. It should be carefully com- 

 pared, at an observed temperature, with the standard of the U. S. Coast and 

 Geodetic Survey, both before and after its use in base measurement. Prefer- 

 ably, the site for the base line should be selected along a railway tangent, 

 as such a location is approximately level, and the railway ties afford an 

 excellent support for the tape. If such a location can not be obtained, it 

 should be selected so as to fill the requirements above mentioned, cleared 



