32 EEPOET OF THE SECRETARY. 



with that of Regnault in the College de France, and that of the ob- 

 servatory of Paris. On his return Professor Guyot made a new series 

 of comparisons with the same Smithsonian standards as before, 

 the results of which proved that no change had taken place during 

 his absence in the two barometers used abroad. It is believed 

 that these comparisons establish a correspondence of the Euro- 

 pean and American standard barometers within the narrow limit 

 of one or two-thousandths of an inch. A large standard is about to 

 be prepared under the .direction of Professor Guyot, to which these 

 determinations will be referred. 



The meteorological branch of the operations of the Institution still 

 continues under the charge of Mr. William Q. Force, to whose habits 

 of order. and scrupulous accuracy the system is much indebted for 

 whatever value it possesses. 



Among the contributions to the meteorological materials of the 

 Institution, presented during the last year, is a series of continuous 

 records of the changes of atmospheric temperature made by a ma- 

 chine invented by Dr. James Lewis, of Mohawk, New York. This 

 instrument consists of an arrangement of a number of brass and iron 

 wires, whose relative contraction and expansion give motion to 

 a metallic point, the several positions of which are marked by a 

 puncture in a fillet of paper, produced by a blow of a hammer 

 moved by clockwork, repeated at regular intervals of fifteen minutes. 

 Although instruments of this kind have been frequently constructed, 

 they have generally not possessed sufficient sensibility to indicate the 

 fitful changes of the atmosphere. Dr. Lewis appears, however, to 

 have been more successful in his invention, and from the results 

 which he has presented to us it would appear that the registers of his 

 self-recording instrument are of considerable value in determining 

 the general law of changes of temperature, especially during the day, 

 and thus furnishing corrections by which the mean temperature of 

 places in the same latitude can be obtained from observations made 

 at only one or more hours of the day. It may be mentioned here 

 that an idea has been very prevalent in this country among observers 

 that to obtain the average temperature of a place, the best times of 

 observation are about sunrise and sunset, and at 2 or 3 o'clock in the 

 afternoon; but as the rising and setting of the sun occurs at different 

 times in different seasons of the year, and as the maximum tempera- 

 ture occurs at different hours in different latitudes, it is best alwaj^s 

 to make the observations as nearly as possible at fixed hours, as, for 



