Bache.] ^'^^ [May 6, 



fore, for the moderate elevation that must be adopted for the bal- 

 loon in order to manage it, we must, with broken surface, accept 

 greater error in delineation than would attach to the same surface 

 if greater elevation of the balloon were permissible. But we should 

 be reconciled to this fact from the consideration that, even were it 

 possible to manage the balloon at the height which would virtually 

 eliminate the error of projection mentioned, the scale of the re- 

 sulting map would be so small as to approach in character the 

 results of a reconnoissance. Another circumstance should recon- 

 cile us to the insuperable fact mentioned, and that is that there are 

 thousands of square miles in our country where, from the very fact 

 that the surface is essentially level, the optical difficulty attaching 

 to moderate elevation for the balloon would not exist. 



Such a survey, by balloon photogrammetry, as that described 

 could be very easily plotted by final process of photographic print- 

 ing. In consequence of the fact that the balloon would be kept at 

 a fixed height throughout a given survey, the scale of the links of 

 the traverse line would be established through the photographic 

 presentment of the length of those links. The scale of those links 

 may also be fixed by the measurement of them on the ground. So 

 the photographic scale and the other scale may be made the same, 

 and therefore they would be made the same. The traverse line 

 having finally been laid down on helios paper, before the paper is 

 sensitized, the paper would then be sensitized, and the photographic 

 plates representing the links of the traverse line would be simul- 

 taneously adjusted upon it along the traverse line as plotted, one 

 scale, as derived from adjusting the balloon at a certain height, and 

 the other scale, virtually the same, as derived from linear measure- 

 ment along the ground, being made to accommodate themselves 

 graphically to each other, thus eliminating error in the resultant 

 map. This resultant map, if the picture of a plane surface, would 

 have but one defect, that of exhibiting minute triangles of blank 

 space where the photographic plates, cut off so as to fit along the 

 links of the plotted traverse line, would necessarily not fill out 

 entirely the delineation of the ground at those points, although 

 otherwise perfecting it elsewhere, from the fact that they would 

 form with one another a continuous series. 



The captive balloon, if used only on days fit for ordinary field 

 work, would occupy a position of almost stable equilibrium, if its 

 power of flotation were sufficient, not only to support the photo- 



