LECTURES. 139 



tion are placed in isiicli a way that they turn their labels towards the 

 interior of the room. 



This arrangement has the great advantage, that when one particular 

 map is looked for it is not necessary to take out the whole parcel 

 to which it belongs, and to search for it among many other maps. 

 Each document can easily be selected by looking over the labels, 

 without disturbing the rest. 



On the other hand^ however, this manner of arrangement, which is 

 observed in nearly all the American chartographical collections, and 

 which is excellent for their particular purposes, offers for ours some 

 great disadvantages. 



First, the maps when they are rolled, and still more so when each 

 roll is put in a separate cylindrical box, as is done for protecting the 

 maps in the arcliives of the United States Coast Survey, take up a 

 much greater space than when the plain sheets in their flat state are 

 laid one over the other. We can easily put in one case of a moderate 

 size a hundred maps, sheet over sheet, while perhaps six times as 

 much space would be required if we rolled them. Besides, the rolling 

 of the maps, the unrolling and flattening them, the troublesome fas- 

 tening of the little bands, &c., have their inconveniences, and the 

 maps must be particularly prepared and strengthened for these often 

 repeated processes. 



But the principal objection is, that the rolling system would be 

 directly against the spirit and tendency of our historical collection : 

 this being destined to show how the maps grew out from each other, 

 it will often happen that a whole series of connected maps is to be 

 consulted. Here it is essential that the chronological order of the 

 maps in every division should always be preserved, which might be 

 difiicult in the process of unrolling, since maps thus managed would 

 always be liable to interfere with one another, and thus get into 

 confusion. 



I am led therefore to the conclusion, that our maps ought to be 

 deposited flat in broad, commodious drawers, one above the other. 

 Labels with numbers and titles may always be added to each of them, 

 in case it should be considered requisite. The drawers will only 

 serve as a receptacle ; for carrying a whole division of maps out of 

 them, and for moving them to the tables for exhibition and back 

 again to the drawers, they may besides be surrounded by a portfolio 

 of pasteboard. 



In no way, however, should our maps be bound up like the sheets 

 of an atlas or a book. They should, in the beginning at any rate, be 

 kept as loose sheets ; because, as has been said, the whole collection 

 must be pervaded by a spirit of progress and growth, and each article 

 be prepared at any moment to cede its place to another newly intro- 

 duced. Every map should also be ready for being transferred from 

 one class into another, and every class for separation into two or three 

 other classes, if the richness of materials in any division should be 

 such as to authorize it. Even the more ancient deposits of our col- 

 lection should be kept, at least for some time, in the same movable 

 state ; because the archives and libraries of Europe might always 

 throw up some old map which had escaped our attention. Sooner or 



