far the results of our measurements up to nearly 14,400 feet 

 seem to justify their use to an altitude of 15,000 feet. 



We will now devote a little space to the aneroid barome- 

 ters, and a few facts that were noticed concerning them dur- 

 ing our stay in Colorado. The best one of these instruments 

 was given to the Palaeontologists when we reached Fort 

 Bridger, and the other was left in our camp while we were 

 in the mountains, and used but little. 



They were taken more for the purpose of giving them a 

 series of severe tests than for any work that we expected of 

 them. One or the other of them was always carried to the 

 top of each mountain we measured in Colorado, having been 

 previously carefully set and compared with the mercurial 

 barometers. In the measurement of the eight mountains upon 

 which they were used they always acted in the same peculiar 

 manner, seldom reaching their true height within half an 

 hour of the time they were on the top ; and on the return 

 having as much hesitation about coming to a standstill. 



There seem to be two great causes of such actions. The 

 first and principal one is the unequal expansion and contrac- 

 tion of the metallic box upon which the measurements 

 depend ; the second of these causes is the mechanical work 

 performed by the same box in the rotation of the index. 



This last difficulty can be, and in fact has been, overcome 

 in that form of aneroid made by Goldschmid, of Zurich; 

 but the first cannot be as easily cured, as it seems to be a 

 defect in the system employed. The metal having once been 

 expanded can never return to the same condition, a change 

 having probably taken place in the situation of its molecular 

 components. 



We give the following as examples of this action of the 

 aneroids ; they are a fair sample of the rest, and not exagger- 

 ated ones : 



Pike's Peak, July $th, 1877. Barometer 2233, at Lake House, two feet above the 



ground. 



