30 Agricultural .Manual 



Three observations of the temperature were made each day, the 

 first being just before sunrise, the second at 3 p. m., and the third 

 just after sunset. 



Beginning about 1890 the thermometers in use were gradually 

 supplanted by self-registering maximum and minimum thermome- 

 ters, the use of which reduced the work to a single observation 

 each day. 



RAIN GAUGES 

 There were five types of rain gauges used in this series, as fol- 

 lows: the Regents gauge (1825-1833); the DeWitt (1834- 

 1848) ; the Smithsonian (1849-1889 I ; thePuertes (1890-1899) ; 



U. S. Weather Bureau Standard (1900 to date). It is not to be 

 understood that all gauges at individual stations were changed 

 when a new type of gauge was adopted, but that the change from 

 one type to another was made gradually as the gauges at indi- 

 vidual stations became unserviceable. In order to determine 

 whether or not there were serious discrepancies in the records due 

 to the use of different types of gauges, the Weather Bureau at 

 Ithaca made a series of comparative observations covering five 

 years, using the types of gauges above mentioned, together with a 

 U. S. Weather Bureau Standard gauge. 



A study of the data thus obtained indicated that the DeWitt 

 gauge, in use from 1834 to 1848, gave results about four per cent 

 *>elow those of the Standard gauge and that the Smithsonian and 

 Fuertes gauges, in use from L8 !'.) to 1899, gave results about three 

 per cent greater than the standard. 



It has not been possible to ascertain when and for what period 

 the different gauges were in use at individual stations, and for 

 that reason an attempt to apply a correction to the records is not 

 considered advisable. Taking the series as a whole, it would seem 

 that the errors arising from the use of the different types of gauges 

 tend, in a large measure, to counterbalance each other and that 

 id. as a whole, may be taken as fairly representing the 

 conditions. 



SNOWFALL MEASUREMENTS 



The most serious error in the records is undoubtedly in con- 

 nection- with the measurements of precipitation in the form of 



