NATIONAL GEOGRAPHIC MAGAZINE. (continued 3)

Every map is a sketch, which is corrected by the geometric location of a greater or less number of points. Assuming entire accuracy in the location of the points, that is, assuming that the errors of location of the points are not perceptible upon the map, the measure of accuracy of the map consists in the number of these geometric locations per unit of surface, per square inch, if you will, of the map. The greater the number of these locations the greater the accuracy of the map, but however numerous they may be the map itself is a sketch, the points located being simply mathematical points. Whatever method be employed for making these geometric locations, the sketching is substantially the same everywhere. The methods of making these locations must differ with the character of the country, as regards the amount and form of its relief, the prevalence of forests and other circumstances. There are two general methods of making the geometric locations used in surveying; one, by triangulation; the other by the measurement of a single direction and a distance, which is the method employed in traverse surveying. In practice, the two methods are often combined with one another. Both methods have been employed in Massachusetts. The fundamental basis of the work was the triangulation which had been carried over the state by the U. S. Coast and Geodetic Survey. By this survey points were located at wide intervals over the state. Besides this there was executed between 1830 and 1840, at the expense of the state, a triangulation known as the "Borden Survey." This located a much larger number of points, but less precisely. The Coast and Geodetic Survey kindly undertook the adjustment of this triangulation to an agreement with its own work, and, as many of the lines were common to the two pieces of work, the locations made by the Borden Survey were by this adjustment greatly strengthened. Even after this work was done, however, there remained considerable areas which were destitute of located points, and it became necessary to supplement it. This was done in part by the Coast and Geodetic Survey and in part by the Geological Survey. By these several agencies upwards of 500 points were made available for the use of the topographers. These are in the main well distributed, furnishing upon each sheet a sufficiency, while upon many the number is greatly in excess of the requirements.

The work of location has been done in different parts of the state by different methods as seemed most applicable to the differing conditions of relief, forest covering and culture. Throughout most of the western part of the state the work was done entirely with the plane table, using the method of intersections as the means of location. Each plane table sheet comprised one-half of an atlas sheet, cut along a parallel of latitude. The plane tabler, starting with three or more locations upon his sheet, furnished by the triangulation, expanded over the sheet a graphic triangulation, locating thereby a considerable number of points, before commencing detailed work. This was done as rapidly as possible consistent with a high degree of precision. The reason for covering the sheet with the graphic triangulation beforehand lay in the necessity for locating a considerable number of points before the sheet had opportunity to become distorted by alternations of moisture and drying. This done, the plane tabler went on with his usual routine of work, locating minor points and sketching the topography in contours. The map was as far as possible completed upon the stations, with the country in view. Elevations were determined as the work progressed, with the vertical circle of the alidade, and minor differences of elevation between points whose height was known were measured by aneroid barometer.

In this work several different forms of plane table have been employed. It was commenced with the large heavy movement designed I believe by the Coast and Geodetic Survey. This, however, was found unnecessarily heavy and cumbersome, and it was discovered that the requisite degree of stability could be obtained with much less weight. For this plane table movement there was soon substituted another form in use in the Coast and Geodetic Survey, which is very much lighter. This was soon improved by taking off the slow motion in azimuth, which was found to be unnecessary, and the addition of more powerful clamps, for the purposing of rendering it more stable. A still more stable form, however, coupled with even less weight, was designed by Mr. W. D. Johnson, of the U. S. G. S. and was immediately adopted. This is substantially a modification of the ball and socket movement. It consists of two cups of large size fitting closely to one another and working within one another in such a way as to allow of the adjustment in level, and the clamping of the level adjustment independently of the azimuth movement, clamps for both level and azimuth adjustments being underneath the instrument. This form is extremely stable, admits of quick adjustment and leveling, and it has been from the time of its invention in general use in this state and elsewhere in the Survey.

In the undulating, forest-covered, region in the southeastern part of the state it was found impracticable to use economically the method of intersections, and resort was had to the traverse method for making locations. In this method, as is well known, one station is located from another by the measurement of a distance and direction, the line of stations being connected at each end either upon stations in the triangulation or upon other lines, while from the stations in these traverse lines, points off the lines are located by intersections, if practicable, or by distance and direction measurement. For this kind of work the plane table, at least such a plane table as is generally in use is an inconvenient instrument. The plane table with the telescopic alidade is too cumbersome an instrument to be carried about and set up as frequently as is necessary in this work. Therefore for this purpose theodolites, fitted with stadia wires and stadia rods, have been used. Distances are measured by the angles subtended by the stadia wires upon the rod, whose divisions are of known length, while the directions are measured by the compass attached to the theodolite, and differences of elevation by spirit level and vertical angles. With this instrument lines were run along all the roads and along the principal streams in this part of the state and from these lines the country lying between them was located and sketched.

In the northeastern and in much of the middle portion of the state a mixed method of work was employed, the plane table being used for carrying on the intersection work wherever it could be done, while by traversing the roads, their details, which could not be obtained by the plane table in this region, were reached. These traverses were platted in the office and the maps drawn from notes and sketches made in the field.

The degree of accuracy of the map depends upon the accuracy of the locations, their number and the uniformity of their distribution. Of their accuracy it is only necessary to state that their errors are not sufficiently large to be appreciable upon the scale of the map, for instance the scale being one inch to a mile, an error of 50 feet in the location of a point would be upon the map but one hundredth of an inch,—a barely appreciable quantity, and it is of course easy to make the locations within this limit. Of the number of locations per unit of map surface I shall give statistics drawn from the full experience of the Survey in this state. The area surveyed by the method of intersections exclusively comprises 3,500 square miles, or about two-fifths of the state. In this area 3,123 stations were occupied with the plane table, or slightly less than one to a square mile, or, measured upon the map, one to a square inch. Besides these, 17,846 points were located in this area by intersections, making, with the occupied stations, a total of 20,969 locations within the area, or 6.2 horizontal locations per square inch. In the same area the heights of 34,893 points were measured, being 10 per square inch. I am expressing these figures in terms of inches of the final map, because it is the map with which we are concerned.

The area surveyed by the traverse method is 2500 sq. miles. In this area 5615 miles of traverse lines were run, being 2.2 linear inches per square inch of the map. In running these lines 46,524 stations were made with the theodolite, being 8.3 per linear mile of traverse and 18.6 per sq. inch of map. The number of measurements of height was 92,561, being 37 to the square inch.

The area surveyed by the mixed method comprised 3000 sq. miles. In this 900 stations were made with the plane table, and from them 3718 points were located by intersection, making altogether 4618 points located with the plane table. In addition to this, 6767 miles of traverse were run, being 2.2 linear miles per square mile of area. In these traverses 31,708 instrumental stations were made, or 4.7 per linear mile and 10.6 per sq. mile. The sum of the plane table stations, locations, and the traverse stations, which makes up the total of horizontal locations in this area, is 36,326, being a total of 12.1 points per sq. inch of map. The number of measurements of height in this area is 67,119, being 22.4 per sq. inch. It will be seen that the number of horizontal locations and of height measurements in the area traversed is much greater than in that surveyed by the intersection method, and it might be inferred that the former work is better controlled than the latter. I do not judge, however, that this is the case, owing to the fact that traverse stations are not of as much value for purposes of location as those by intersection. The latter are selected points. The former are not selected points, but on the contrary, a large proportion of them are located simply for carrying forward the line and are of no further service, and very few of them are such as would be fitted for the purpose of controlling areas.

Within the area surveyed by traverse nearly every mile of road has been run. With the exception of those in the cities, nearly every house and every church in the commonwealth has been located, either by intersection with the plane table or by traverse.

The organization of the surveying parties has been of the simplest character. Plane table work has been carried on by one man with an assistant, the latter doing little more than attend the plane tabler and assist him in carrying the instruments. Each of these little plane table parties was furnished with a horse and buggy for transportation. The organization for traverse work has been equally simple, consisting of a traverse man and a rodman. As a horse and buggy would be an impediment in this work, this feature of the outfit has been omitted. In the mixed work the traverse men have been under the immediate control of the plane tabler, so that their movements have been directed by him in detail. The average output per working day of the plane tabler has been for the whole survey 3.1 sq. miles, and of the traverse man 2.8 sq. miles, and, as the expenses of the former have been slightly greater than those of the latter, the cost per square mile of the two methods of work has been substantially the same.

The average cost per square mile of the survey of the State has been a trifle less than $13. This includes the salaries of all men engaged upon the work during the field season, their traveling, subsistence and all other expenses, the salaries of the men engaged in drawing the maps in the office, the cost of supervision and of disbursement,—in short all expenses of whatever character, incurred in the production of the map.

PROCEEDINGS

OF THE

NATIONAL GEOGRAPHIC SOCIETY.

ABSTRACT OF MINUTES.

First Regular Meeting, Feb. 17, 1888.—Held in the Law Lecture room of Columbian University, the president, Mr. Hubbard, in the chair.

The president delivered an inaugural address.

Major J. W. Powell lectured on the Physiography of the United States.

Second Regular Meeting, March 3, 1888.—Held in the Law Lecture room of the Columbian University, vice-president Bartlett in the chair.

Paper: Patagonia, by Mr. W. E. Curtiss.

Third Regular Meeting, March 17, 1888.—Held in the Assembly Hall of the Cosmos Club, the president, Mr. Hubbard, in the chair.

Paper: Physical Geography of the Sea, by Commander J. R. Bartlett.

Fourth Regular Meeting, March 31, 1888.—Held in the Assembly Hall of the Cosmos Club, the president, Mr. Hubbard, in the chair.

Discussion was had on the proposed Physical Atlas of the United States, participated in by Messrs. Gannett, Gilbert, Ogden, Greely, Marcus Baker, Willis, Bartlett, Merriam, Ward, Henshaw and Abbe.

Fifth Regular Meeting, April 13, 1888.—Held in the Assembly Hall of the Cosmos Club, vice-president Merriam in the chair.

The discussion of the proposed Physical Atlas of the United States was continued, and was participated in by Messrs. Marcus Baker, Greely, Willis, Cosmos Mindeleff, Gilbert Thompson, Kenaston, Gannett and Van Deman.

Paper: The Survey of the Coast, by Mr. Herbert G. Ogden.—(Published in Vol. 1, No. 1, "National Geographic Magazine.")

Sixth Regular Meeting, April 27, 1888.—Held in the Assembly Hall of the Cosmos Club, the president, Mr. Hubbard, in the chair.

Papers: The Great Storm of March 11–14, 1888, by Gen. A. W. Greely and Mr. Everett Hayden.—(Published in Vol. 1, No. 1, "National Geographic Magazine.")

Geographic Methods in Geologic Investigation, by Prof. W. M. Davis.—(Published in Vol. 1, No. 1, "National Geographic Magazine.")

Seventh Regular Meeting, May 11, 1888.—Held in the Assembly Hall of the Cosmos Club, vice-president Merriam in the chair.

Papers: The Survey and Map of Massachusetts, by Mr. Henry Gannett.—(Published in Vol. 1, No. 1, "National Geographic Magazine.")

Graphic Triangulation, by Mr. W. D. Johnson.

Eighth Regular Meeting, May 25, 1888.—Held in the Assembly Hall of the Cosmos Club, vice-president Merriam in the chair.

Papers: The Classification of Geographic Forms by Genesis, by Mr. W. J. McGee.—(Published in Vol. 1, No. 1, "National Geographic Magazine.")

The Classification of Topographic Forms, by Mr. G. K. Gilbert.

The North Winds of California, by Mr. Gilbert Thompson.