516 THE POPULAR SCIENCE MONTHLY. 



of which the earlier are more or less rough ad unconnected, and even 

 some of later dates cannot be entirely relied upon. 



Running surveys, more or less in detail, are generally the precur- 

 sors of the more strict geodetic survey, but, in order to answer the 

 wants of navigation, these should always be based upon a triangula- 

 tion between natural landmarks, checked at reasonable distances by 

 A^ery careful shore observations for latitude and longitude, and the 

 latter carried directly from a central position to the most prominent 

 points of the thus surveyed area and back again, and the central posi- 

 tion connected in the same manner with the nearest satisfactorily 

 determined position, to which the longitudes of that locality are gen- 

 erally referred. 



The telegrajjh-cables which already connect many of the most 

 important places will soon gird the globe in several belts, and will 

 afford the means for ascertaining great meridional distances with 

 almost absolute correctness. There will thus be furnished a great 

 number of primar}^ positions from which the longitude may be carried 

 in coordinate lines to secondary places. In this manner a network 

 of points spread over the globe will be attained, corresponding to the 

 primary and second triangle points of great geodetical operations. 



The completeness and correctness of a running survey depend 

 upon the time devoted to it and the difficulties encountered ; frequent- 

 ly the coast-line is only traced in from point to point, or from the 

 shore-ends of the lines of soundings by the eye ; the points of land, 

 however, especially the salient ones, should always be fixed by angles 

 to or from the established landmarks, as should also all outlying 

 dangers and all features bearing directly upon or assisting navigation. 



The surveys of harbors and anchorages should be as complete as 

 possible ; if time jjerraits, beacons should be erected for triangulation, 

 and the plane-table employed for obtaining the shore-line. The parts 

 of the latter which are merely traced in approximately should be dis- 

 tinguished on the chart by a broken line. 



The soundings should always be numerous enough to show the 

 configuration of the bottom of harbors, and off a seacoast the gradual 

 rise from great depths to the shore, islands, and banks, so that the 

 characteristic curves of the depths may be shown with precision on 

 the charts ; for harbors generally the one, two, three, and five fathom 

 curves are marked ; on coast-charts, those of three, five, ten, twenty, 

 fifty, and one hundred fathoms. 



When sounding from a vessel in motion or from a boat, the lead 

 should be tried at intervals, even when it is anticipated that the bot- 

 tom will not be readied, not only on account of the possibility of the 

 discovery of a sudden elevation, but for the purpose of placing the 

 negative soundings on the chart, which show conclusively the absence 

 of danger and that the ground has been examined. 



For such negative soundings, as pauch line should be used as the 



