778 GEOGRAPHICAL LATITUDE. 



to say something of the work in connection with measuring a base line 

 and fixing and measuring the triangles connected therewith. In the 

 first place, choosing the ground for the entire survey is no easy mat- 

 ter. Two points nearly or quite on the same meridian should be sought, 

 which are favorable to the astronomical observations necessary to de- 

 termine their latitude ; near the course of the line there should be a 

 convenient plane, as flat as possible, advantageous for running the base 

 line in a direction favorable for connecting it with the main series of 

 triangles, while the tract as a whole should be of such a nature as to 

 permit the making of triangles as large as possible, L e., furnishing 

 high points at long distances with uninterrupted view, with all the an- 

 gles as large as possible, for the measurement of very acute angles is 

 more liable to error than that of large ones.^ The situation of the base 

 line once selected, its end points are marked with the utmost care by 

 contrivances which vary according to the ideas of the surveyor. In 

 order to keep the true direction with perfect accuracy while measur- 

 ing, the line is staked out with the utmost care, wooden stakes being 

 driven into the earth at short distances, and nails driven into the top 

 of the stakes to mark the exact point where the line passes. Then fol- 

 lows the actual work of measuring. That this may be accomplished 

 with the most extreme accuracy, a great variety of rules have been con- 

 structed from time to time, the standard of length generally being the 

 toise of Paris, or of Peru, as it was called after the completion of the 

 Peruvian survey. The first care is to procure a rule as nearly as possi- 

 ble just so long or double so lonrj as the standard and determine its 

 absolute length. Then, since the substance thereof (of whatever sort) 

 is liable to variation of volume dependent on change of temperature, ex- 

 periments must be made to establish the amount of this variation, and 

 a thermometer is so attached as to give the temperature of the instru- 

 ment. There must be an attachment to the rule which enables it to be 

 placed at a perfect level, or if not, then to measure the amount of the 

 declination. An arrangement for placing it firmly on the ground and 

 at the same time for changing its level must also be thought out and 

 constructed. Furthermore, since it is practically impossible to place 

 one heavy rule against another without displacing more or less the one 

 already m situ, ingenuity has been taxed to invent methods by which 

 the main rules could be placed in line without touching and the small 

 intervals be accurately measured. This difficulty can be obviated by 

 the application of the princii)leof wheel measurement, by which, a per- 

 fectly smooth way being constructed between the two points whose dis- 

 tance is to be measured, a cylinder, of special construction, the length 



• Bongner (Figure de la terre., p. 79) says in this connection : " Ces erreurs, quoique lea 

 memes, produiront cependant ensnite diff^rens effets, selon que les angles eeront plus 

 on moins grands; une minute apporte beaucoup plusde diffi^rence dans le sinus d'lin 

 petit angle que dans le sinus d'un grand ; et les c6t^s qu'oii calcule par le moyeu 

 des triangles, doivent etre sujets a la niAnie errenr que ces sinus, puisqu'ils cluingent 

 daus le meuie rapport." See also pp. 83, 85, 88. 



