178 EXPLORATIONS ACROSS TEE GREAT BASLN OF UTAH. 



It is scarcely necessary to remark that, in the application of these corrections, 

 proper discretion is required on the part of the computer, and that the tables are merely 

 intended to help him. While the variations are smaller on a clouded and rainy day, 

 not favorable for radiation, they are larger on a clear day, and much depends upon 

 local circumstances, and the direction and force of the winds, &c. When the successive 

 camps and minor stations are in the same valley, or do not differ much in altitude and 

 physical relations, the mean temperatures may be determined with great precision; 

 but where the altitude and relative position of the stations vary much, as they did on 

 our survey in Utah, from low, arid valleys or scorched slopes to narrow canons or high 

 mountain summits, it is very difficult to determine the mean temperature of the day 

 from one or a few observations, the more so because the hour of the maximum tem- 

 perature also changes according to the relative situation of the stations. 



SELECTION OF A FIXED STATION. 



After all these corrections had been applied, the observations were ready for com- 

 putation. The tables of Prof. A. Guyot, based on La Place's formula, were used for 

 this purpose. Next the question arose what should be taken as the lower or fixed 

 station for the calculation of the altitudes in Utah. As most of them are considerably 

 high, between 5,000 and 8,000 feet, the air-temperature appears as an important ele- 

 ment in the computation. A difference of 1 degree in the temperature changes the 

 result 1 foot for every 900 feet of the height. By taking the sea-level as the lower 

 station, with a comparatively high mean temperature, this element appears to exercise 

 an unduly great influence on the result, after all the corrections nave been applied, 

 which, if fully answering the purpose, would require the mean temperature of the 

 year to be used, not the one, generally much higher, of the day of the observation. 

 Camp Floyd was an elevated inland station, for which the mean reading of the barom- 

 eter and thermometer could be ascertained, and the altitude of which could, therefore, 

 be determined satisfactorily. By taking Camp Floyd as the lower station, I decreased 

 in a o-reat measure the influence of the temperature in the computations, and all errors 

 arising from that source. The altitudes of all places not very far from Camp Floyd 

 were certainly obtained much more correctly in this way, and I believe also most of 

 the others; at least I obtained by this method results which agreed very satisfactorily 

 in several cases when observations, taken at different times, controlled each other, 

 while the use of the sea-level as the lower station would mostly have given greater 

 discrepancies. 



It might be urged that, in case the altitude of Camp Floyd was not correctly 

 determined, this error would be propagated by assuming it as the fixed station. The 

 error in the altitude might originate from various causes: Firstly. The values assumed 

 as mean readings of barometer and thermometer at the level of the sea might not be 

 those best adapted for the special purpose; then the computer would introduce the 

 same causes of error into the other calculations, and the results would be obtained 

 even more uniform on assuming an intermediate station for references. Secondly. The 

 mean reading of barometer and thermometer, as given for Camp Floyd, might not be 

 absolutely correct This error cannot be great. If the values did not correspond to 



