46 Journal op the Mitchell Society [Sepfeniher 



the plus distance back to the preceding station. This method avoids 

 measuring the tangent distance from the P. I., in order to locate the 

 P. C. 



It should be noted that the curve can be located b.y starting at 

 the P. C. instead of the P. T., the deflection to the first point being 

 the same as that to the 9th jioint as given by Table 1, etc. 



If part of the curve is not visible from the P. I., say that beyond 



point 6, then the instrument may be moved to point 6 and the reaiain- 



ing points located by deflections from a preceding chord. Suppose a 



backsight is taken to point 2, vernier reading 0°, then after reversing 



' 5A° A<» 



the telescope the proper deflection to locate point 7 is = — , 



etc. 2 X 10 4 



For very long curves in woods and at places where the P. I. is 

 not accessible, the well-known deflection method should be used, that 

 is, the instrument is moved to the P. C, and intermediate points on 

 the curve. 



During the past summer the writer was in charge of a party that 

 surveyed 22 miles of federal-aid highways in hilly and mountainous 

 country, and the following facts were observed: (1) not a single case 

 of inaccessible P. I. occurred; (2) along 86 per cent of the curves the 

 P. I. was visible throughout; (3) in only 42 per cent of the curves 

 was the P. T. visible from the P. C. This means that 86 per cent of 

 the curves could have been laid out completely with the instrument 

 set only once (at the P. I.), whereas 58 per cent of them actually re- 

 quired the instrument to ])e set up three times. The first 11 miles was 

 in fairly open country along the general direction of an old road. 

 Here 96 per cent of the curves were visible throughout from the P. I., 

 but 70 per cent of the P. T. points were not visible from the P. C. 

 The other 11 miles of the survey was partly in a dense forest and not 

 along an old road. 



Aside from the time saved in not having to move the instrument, 

 another step in the usual operation of laying out a curve is avoided, 

 and that is, the tangent distance is not measured from the P. I., in 

 order to locate the P. C. Curves laid out by the usual method begin 

 and end with subchords of unequal length. This makes the deflections 

 rather tedious to compute. The errors are eunuilative, and tlie writer 

 has seen the best of transitmen waste time in trying to find the little 

 error that prevented the final check. 



