to run through the printing presses a sheet with 27 colors, averaging about one color a day-the changes in humidity may be very great and are known to have caused a change of one-eighth of an inch in the length of a sheet. To overcome this trouble as far as possible the room in which the partly printed sheets are stored is kept at uniform temperature, the windows being closed the year round. A humidityrecording instrument is also kept in the room as an indicator of conditions.

The structure and columnar sections are usually prepared by the draftsmen of this section of the Survey, who copy the outlines of the draft submitted by the authors, but use a definite system of lithologic symbols to represent the kinds of rock, in this way obtaining uniformity of style and symbols. Thus, dots represent sand, sandstone, and quartzite, the coarseness being indicated by the size of the dots; parallel dashes represent clay, shale, and slate; parallel lines united here and there by short perpendicular lines to form rectangles represent limestone; irregularly placed dashes represent igneous rocks. For rocks intermediate between these types the appropriate patterns are combined.

The structure sections are engraved and transferred to blank strips left for them on the structure-section sheet, the outcrop of the formations on the sections corresponding with the boundaries on the map along the lines of the sections. Small strips of the lithologic symbols used in the sections are engraved in the corresponding blocks of the legend. Sheets are then colored for the sections similar to those prepared for the maps, and the color proofs are read. A portion of one of the geologic maps with section is shown in Pl. II (p. 30).

The cover of the folio not only serves the purpose of protecting the contained sheets, but bears on its front page an index map showing the location of the area mapped, and on its inside pages a description of the topographic map and a brief outline of geologic processes, intended to prepare the lay reader to understand the more detailed and somewhat technical description within.

The first of the final series of folios was completed in 1894. In the succeeding ten years 106 folios have been published, an average of over ten a year. Three of the folios each include maps representing four contiguous areas, and eight folios embrace two such contiguous areas each, so that in labor and time they represent a correspondingly larger number of ordinary folios. Seven folios contain special detailed maps of mining regions, and two more are in process of publication; eleven have special artesian-water maps. Three folios embracing large cities (New York, Chicago, and Washington) have been published.

A table of the published folios is given in the list of Survey publications.

Section of Topographic Maps.

When reading proofs, the author of a map which he has prepared without special editorial assistance, be he surveyor or compiler, like the author of a text, is apt to discover that he has omitted important data and that his manuscript is more or less indefinite, ambiguous, and inaccurate. Correction of engraved plates is more difficult and expensive than correction of ordinary type. Moreover, a cartographic output so large as that of the Geological Survey should exhibit uniformity and system as well as clearness and accuracy. For these reasons all topographic maps to be published by the Survey are critically examined by the editor of topographic maps before they are placed in the hands of the engraver.

At the beginning the Survey established a standard form and size for its atlas sheets. In a general way it prescribed the information which the sheets should bear and the manner of its expression, but the topographers of the small force exercised wide discretion as to details. As time went on and the organization grew in size and complexity the necessity increased for prescribing in greater detail the classes of facts which the maps should portray, for greater differentiation in the symbolism, and for closer adherence to the standards thus established. At first such editing as the topographic maps received was given them by the chiefs of the various sections of the topographic branch. Naturally there was lack of uniformity, and it was not long until the need of systematic editing became apparent. This resulted in 1894 in the appointment of an editor of topographic maps, to whom all manuscript atlas sheets were submitted before they were transmitted to the engraving division. Gradually, with more careful and systematic editing and proof reading, this section has grown until its personnel includes an editor and five assistants.

The routine at present is as follows: The topographic section chief transmits the manuscript topographic map, through the Director, to the editor of topographic maps, who becomes the custodian of such manuscript both before and after publication. In his office the manuscript receives systematic examination, after which it is returned to the chief of the topographic section for revision and the supply of further information where necessary. The editor also proofrea prints from the copperplates, comparing them in detail with the matscript, and transmits the final proof to the topographic branch for revision and approval before printing.

In order that nothing shall be overlooked in the editing of the topographic atlas sheets, a printed schedule is followed, which shows the subjects of examination ranged in the order of procedure. The examiner also provides himself with a photograph of the manuscript, on the scale of publication, on which he marks with red ink all changes, corrections, explanatory notes, and notes of instruction to the

engravers.

He notes, first, whether the manuscript atlas sheet conforms to the standard in form, dimensions, scale, contents, and symbols. On the margins of the atlas sheet are printed certain explanatory names, notes, figures, legends, and diagrams, each of which must have its prescribed form and position. The matching of the sheet in hand with the adjoining sheets is next scrutinized in detail. When two or more atlas sheets of adjoining quadrangles are placed on a flat surface with their corresponding projection lines in contact they should form a continuous map without overlapping. Each line of the map extending from one sheet across the other should be continuous, and on each the same natural feature, public work, or political division should presumably bear the same name.

The editor rarely has available the data necessary to enable him to form a judgment as to the accuracy of the map in respect to topography, drainage, and culture. That would require field inspection. Often, however, by comparison with existing maps of the locality and other publications, he is able to criticise its correctness in some particulars and to detect the omission of certain public works which should be shown. By careful inspection of the drawing he may sometimes detect details of topography or drainage apparently improbable or impossible. The examination of these features of the map, however, is mainly for the purposes of seeing that the map is consistent within itself, of improving the placing and distribution of contour figures if advisable, and of interpreting in advance for the engravers details that might be obscure or doubtful.

For the verification of post-offices, railroads, and railway stations comparison is made with the United States post-route maps and Postal Guide, with railway guides and railway time-tables, and with other available sources of information. The United States land-survey system of township, range, and section lines and numbers, and to some extent the boundaries and names of land grants and reservations, are verified by comparison with the maps and plats in the General Land Office. The names and boundaries of Indian reservations, military reservations, forest reserves, and national-park reservations must be shown, and are verified by comparison with the Presidential proclamations establishing them and with such later ones as modify their boundaries.

The sources of information concerning the names and boundaries of the civil divisions include such official publications as reports and maps issued by boundary surveys and Census reports, as well as county atlases and maps and gazetteers.

The geographic nomenclature receives special attention. The editor sees that none of the prescribed nomenclature has been omitted and that all names are properly applied and correctly spelled. Then the placing of the names is examined, and such changes as will improve the appearance of the map and add to its clearness are indicated. When

one compares a number of maps of the same region, made at different dates covering a considerable period of time, he will find that the same geographic name appears in variant forms and that different names are applied to the same feature. Concerning many of these, it will be found that, while the earlier maps display divergent usage, it is manifest from the more modern that usage has concentrated on some particular form or name. In other cases no evidence of an established usage will be found. The editor seeks to verify every geographic name. The various authorities available for this purpose are the maps, charts, and publications of the different Government surveys and bureaus, past and present, reports of exploring expeditions, railroad surveys, geologic reports, maps and reports of State geologic surveys and mining bureaus, local histories, reports of local historical societies, etc. When uncertainty arises as to the name of a feature or as to the form which the name should take, the different names and variant forms are compiled and the authorities cited. The origin or meaning of the name is sought and recorded, information concerning local usage is obtained through correspondence with local officials and residents, and the whole is referred to the United States Board on Geographic Names for a decision. The names of many minor geographic features are published for the first time on these atlas sheets. No effort is made to verify such a name unless by reason of its unusual form or other circumstance it excites distrust; then correspondence is opened to determine its correctness. Again, a natural feature of such size and importance as to justify the assumption that it has a name may be unnamed on the atlas sheet. Correspondence is the means by which this is determined.

An exhibit of the information gathered concerning a geographic name recently investigated in editing an atlas sheet will serve to illustrate the method pursued and the research necessary before a decision can be made intelligently. It also illustrates one phase of cooperation between the Survey and the Board on Geographic Names.

A cape on the eastern shore of Penobscot Bay, in Hancock County, Me., was named on the manuscript "Rosier;" and a post-office near by, "Cape Rosier." It took but little investigation to discover that the various maps of the region were not in accord as to the spelling of the name. Some spelled it Rosier, while others gave the form Rozier. A more searching examination and compilation of authorities resulted as follows:

ROSIER.-U. S. Geological Survey manuscript Castine sheet, 1900.

U. S. Coast and Geodetic Survey charts 104 and 310.

U. S. Coast and Geodetic Survey, Atlantic Coast Pilot, Part I, pp. 84, 85, 87, etc.

U. S. Light-House Board, List of Beacons, Buoys, and Day Marks, 1900. ROZIER.-U. S. post-route map and Postal Guide (current).

Varney's Gazetteer of Maine, 1886, pp. 137, 138.
Colby's Atlas of Maine, 1885.

It was also learned that the cape was named in honor of a companion of Weymouth and the chronicler of his expedition to the coast of Maine in 1605. A search for his name was successful, as follows: JAMES ROSIER.-Ancient Dominions of Maine, Sewall, 1859, pp. 35, 36, 60.

A Brief History of Maine, in Colby's Atlas, Dr. W. C. Lapham, 1885.

Narrative and Critical History of America, Winsor, Vol. III, pp. 81 and 191.

JAMES ROZIER.-Varney's Gazetteer of Maine, 1886, p. 137.

Correspondence with local authorities elicited the information that local usuage is very evenly divided between Rosier and Rozier; and one of the parties addressed cited the following additional authorities: ROSIER.-Walling's Map of Maine, 1860.

Bangor Historical Society Magazine, Vol. II, No. 11.

ROZIER.-Maine Register.

JAMES ROSIER.-Bangor Historical Society Magazine, Vol. V, No. 12, p. 222.
Abbott's History of Maine, pp. 33, 35, 40, etc.

Maine Historical Society, series 2, vol. 2, index; 21 references.

This information was properly scheduled and submitted to the Board on Geographic Names for a decision. The Board adopted Rosier as the name of the cape and Cape Rosier as the name of the post-office. The special and general maps made by the Survey receive a similar examination before engraving, and the necessary proof reading before publication. All of the maps intended to illustrate the various publications of the Survey in book form, whether made from its own. surveys or compiled, are examined by the editor of topographic maps, with special reference to the geographic names thereon. During the calendar year 1903 there were 223 of these maps examined, this work involving the verification of about 35,000 names, in addition to other cultural and topographic features.

In the beginning the proof reading of prints from the engraved plates was done by various persons in the intervals of other work. On the appointment of an editor of topographic maps this work was placed under his supervision and has since been performed by a corps of trained proof readers. When the engraving of an atlas sheet has been completed a separate print is made from each of the three plates, and each of these is in turn compared, detail by detail, with the manuscript map. Errors and omissions are marked on the proofs, which are then returned to the engravers for correction of the plates. After correction another set of plate prints is made. On these the proof reader notes errors that may have been overlooked in the first proof reading or in the first correction of the plates. These also are returned to the engravers and the plates are again corrected. The engraving is then lithographed and proofs are printed with the three colors combined on one sheet. The combined proof is examined in detail by the proof readers, to see that the three colors are properly registered-that is,