24 INSTRUCTIONS TO MARINE METEOROLOGICAL OBSERVERS 



Form No. 1202 — ISIarine. 



U. S. Weather Bureau Barometer Comparison Card 



Vessel, Amer. S. S. Argonne. Captain C. P. Snyder. 



In port of San Francisco. Observer, G. A. Bryan. 



W. B. List Barometer No. 3430. Mercurial or Aneroid? Mercurial. 



Addi-ess tag to U. S. Transport Service, San Francisco. 



The observer is requested to make at least three barometer readings in order tliat a 

 reliable correction may be obtained. In United States and Canadian ports please make 

 the readings on three successive days, at 7 : 30 a. m., T5th meridian time, or if in a 

 United States port for a shorter period, make three i-eadings as many hours apart as 

 possible. In European ports make the readings at 7 a. m., Greenwich time. 



A tag will be furnished showing the instrumental error of the ship's barometer if the 

 range between the largest and smallest individual corrections in column 4 does not exceed 

 0.10 inch (2.0 millimeters, or 3 millibars). 



Sept. 17 

 Sept. 18 

 Sept. 19 



Time (local) 



7:30 a. m 

 7:30 a. m 

 7:30 a. m 



1 



Ship's ba- 

 rometer 

 (as r^ad 

 off) 



29.83 

 30.06 

 30.16 



Attached 

 thermom- 

 eter 



75 

 67 

 66.5 



Observers will leave these 

 columns blank 



29.68 

 29.94 

 30.04 



29.71 

 30.00 

 30.07 



Correc- 

 tion 



+.03 

 +.06 

 +.03 



In United States ports no postage is required; in foreign ports hand to United States consul. 



Change in haronieteis. — Ordinarily only a barometer bearing an 

 identification number assigned by the Weather Bureau should be 

 used in the regular meteorological work for the bureau. This number 

 should be stated in every "Weather Report and on every Barometer 

 Comparison Card. If, for any reason, another barometer is used, the 

 fact should be made clear in order that no mistake may be made in 

 applying corrections. If a new barometer is used comparative read- 

 ings should be furnished with tlie first report, or as soon thereafter 

 as possible. 



THERMOMETERS 



Thermometers should be of reliable manufacture and should have 

 the graduations etched upon the glass stem. For some years after 

 a thermometer has been manufactured the glass in the bulb con- 

 tracts and, notwithstanding that a certain period of "curing" is al- 

 lowed by good manufacturers, this contraction not infrequently re- 

 sults in appreciable error in the readings. On this account it is 

 necessary to compare the readings of a thermometer with those of a 

 standard instrument to ascertain if any error exists and, if so, the 

 proper correction to be applied to the readings. Ordinarily no 

 thermometer should be employed the indications of which at any 

 point on the scale differ by more than 1° from the true temperature 

 given by a standard instrument. 



Use of thermometers in meteorological observations. — The prin- 

 cipal use of thermometers in meteorological observations is for the 

 purpose of determining the temperature and humidity of the free 

 air. The actual temperature of the air is obtained from the reading 

 of an ordinary thermometer. Tlie humidity is determined by meas- 

 uring the so-called temperature of evaporation and comparing this 



