502 GEOLOGICAL SURVEY OF THE TEREITORIES. 



index error was found to be — .013'', having changed in its journey of 

 seven thousand miles only .003 of an inch, which represents about 3 

 feet in elevation. We made use of small aneroid barometers for taking 

 elevations on the road, and also those of mountains. The results ob- 

 tained were quite satisfactory. But without a mercurial barometer as a 

 check they are almost worthless. Their structure is so delicate as to 

 render them liable to i:)ermanent injar^^ by a sudden fall, or by the con- 

 stant jolting to which they are subjected when carried upon horseback. 

 It was only with the utmost care that the one which was my constant 

 companion was kept in good condition. Another fact noted was the 

 sluggish return of the index to the correct reading, after having been 

 taken to an elevation of several thousand feet. The thermometers used 

 were those manufactured by James Green, and were small-sized Fahren- 

 heits. Two of these were fitted to be used as wet and dry bulb ther- 

 mometers, and furnished much interesting material for the record. 



The distances given in the table were estimated from odometer read- 

 ings, taken by Mr. A. Schonborn, of the expedition. 



To the officers in charge of the U. S. Medical Museum, Washington,D. 0., 

 we are indebted for much valuable assistance in the comparison of the 

 instruments with standards. We have made the suitable corrections in 

 accordance with their advice. 



The abbreviations used in the tables are those customary for the time 

 of day, for the points of the compass, and for the clouds. 



In the column for the force of the wind, indicates a perfect calm ; 

 1 indicates a light air j 2 indicates a i'>leasant breeze j 3 indicates a fresh 

 breeze ; 4 indicates a stiff breeze ; 5 indicates a moderate blow ; 6 in- 

 dicates a light gale j 7 indicates a hard gale ; 8 indicates a very heavy 

 gale ; 9 indicates a very great storm ; 10 indicates a hurricane or tor- 

 nado. 



In the column for the amount of clear sky, indicates a sky totally 

 obscured by clouds ; 10 indicates a sky perfectly free from clouds. The 

 ifitermediate numbers indicate the relative amount of clouds and clear 

 sky. 



The barometric readings have been corrected for index error, and 

 reduced to 32^ Fahrenheit. 



In the calculations for elevation, Loomis's tables, as given in his "In- 

 troduction to Practical Astronomy," have been employed. The morn- 

 ing observations at 6 o'clock, or reduced to that hour, and the evening 

 observations at 9 o'clock, or reduced to that hour, have been preferred. 



