12 MAI'S ACCOMPANYING THE REPORT. 



travelled were measured by an odometer, until the wheels were necessarily abandoned among 

 the Cascade mountains ; and then carefully estimated from the time and supposed rate of travel. 

 The courses were determined by prismatic compasses. The latitudes of a large majority of the 

 camps were fixed with considerable accuracy by astronomical observations. Several camps 

 before camp 17 were connected with San Francisco by chronometric differences, and the longi- 

 tude thus approximately determined. An unfortunate accident, in Canoe Creek valley, however, 

 rendered the chronometers worthless for this purpose during the remainder of the survey, and 

 compelled us to depend upon our courses and distances, checked by the latitudes of the camps, 

 and by a system of triangulation among the prominent mountain peaks near the trail. The 

 assumed longitudes of a few important points upon the route seem to require particular ex- 

 planation. 



As Fort Beading was the point from which we started to leave the settlements, great care 

 has been taken to determine its longitude as correctly as possible. Col. J. C. Fremont, on his 

 map of California and Oregon, places the point of Cow Creek, upon which the fort is now 

 situated, in Long. 122° 6' 50" west from Greenwich. On the Land Office map of 1855, it is 

 placed in Long. 122° 11' 9". On the map of Lieut. E. 0. Beckwith, 3d artillery, illustrating 

 his exploration for a Pacific railroad route near the 41st parallel of north latitude, it is placed 

 in Long. 122° 5' 8". The four chronometers used on our survey apparently preserved their 

 rates unchanged during our march up the Sacramento valley, as they all agreed very well with 

 each other. The longitude of the fort, determined by their mean corrected difference from local 

 time, was 122° 10' 50". As this differs only three-tenths of a mile from that given by the Land 

 Office map, it has been adopted as correct. It places the fort 3.5 miles west of Col. Fremont's 

 location, and 5 miles west of that of Lieut. Beckwith. 



The following method has been adopted to fix the longitude near the northern terminus of the 

 survey. The longitude of Salem has been determined with considerable care, under the direc- 

 tion of the surveyor general of the Territory, both by astronomical observations and by 

 measuring a line to the coast, and thus comparing the result with the work of the United States 

 Coast Survey. It is 122° 5o' 43" west from Greenwich, as I was informed, when at Salem, by 

 Mr. Hervey Gordon, deputy surveyor. He also told me that Mount Hood and Mount Jefferson 

 had been carefully located by bearings taken from well determined points with the solar compass. 

 I therefore made a preliminary plot of the northern portion of our survey, based upon the Land 

 Office positions of these peaks as fixed points. As over fifty bearings had been taken to each 

 mountain, many of which were from points where the latitude was astronomically determined, 

 I was enabled to slightly correct the relative position of the two peaks. The map was next 

 replotted with respect to these new positions. The result was highly satisfactory, as the compass 

 work fitted admirably, and the longitudes of two points in Des Chutes valley, determined by 

 Col. Fremont in 1843, by observing the occultations of Jupiter's satellites, were almost precisely 

 the same as those of the corresponding points on the plot. It is thought that this coincidence 

 renders it very improbable that any important error in longitude has been made. 



The latitude of Fort Dalles was astronomically determined, and numerous bearings upon 

 Mount Hood and the neighboring peaks enabled me to fix its longitude very closely. It was 120° 

 58' 30". This location is about three miles west of that found by Col. Fremont, by observing 

 an emersion of Jupiter's second satellite, on November 5, 1843. He afterwards observed the 

 emersion of Jupiter's third satellite, on November 20, 1843, at the same spot, and published 

 the data obtained, without, however, giving the deduced longitude. I find, by computation, 



