INSTRUCTIONS TO MARINE METEOROLOGICAL OBSERVERS 21 



changes as are indicated by the attached cohimn of mercury, dur- 

 ing the period in which the jar is being exhausted of air. 



Aneroids suspected of temperature trouble may likewise be tested, 

 and shipmasters are cordially invited to avail themselves of the op- 

 portunity of having their barometers given the benefit of this test, 

 which may be had free of charge at the following Weather Bureau 

 port offices : 



Boston, New York, Philadelphia, Baltimore, New Orleans, Port 

 Arthur, San Pedro, San Francisco, Portland, and Seattle. 



Special temperature-testing boxes have been devised, and are in 

 use at most of the foregoing-named stations. Each box contains 

 a copper tank for holding ice-salt mixture, and is also equipped 

 with a thermometer and with 50-watt and 100-watt electric light 

 bulbs. By means of these cooling and heating arrangements, box 

 temperatures as low as 40° F. and as high as 110° F. may be arti- 

 ficially produced. This assures a good thermal range to which the 

 aneroids may be exposed at 10° to 20° intervals of temperature 

 change during the period of the test. 



In cold weather a partial test at high and low temperatures may 

 be secured by obtaining the reaction of the aneroid at the arti- 

 ficially heated room temperature, then, after exposing the instru- 

 ment to the chilled outer air until thoroughly cooled to this outer 

 air temperature, again comparing the barometer with standard at 

 the same elevation. 



If the shipmaster cannot avail himself of the opportunity to 

 have his aneroid tested for temperature at a Weather Bureau office, 

 the instrument division of the Bureau suggests a test method that 

 may be given readily to an aneroid on shipboard. Choosing the 

 early morning of a fair day, simply carry the barometer to the cold- 

 storage room of the ship, being careful to hang it in the same rela- 

 tive position as on the bridge, and allow an hour or more for all 

 parts of the instrument to reach the same temperature. With due 

 regard to the actual change in the barometric pressure, and with 

 allowance for a rise of 0.01 inch for each 10 feet change to a lower 

 level, the effect due to change in temperature of the instrument 

 should not exceed 0.02 to 0.04 inch in a well-compensated aneroid. 

 A similar test for high temperature may be made by carrying the 

 barometer to the engine room. The test as indicated should be made 

 during a period of practically stationary atmospheric pressure, and 

 not during a period of sharp barometer change when the errors, 

 if any, due to temperature would be overshadowed by the natural 

 pressure rise or fall. If a second barometer is available, allowance 

 through comparisons may readily be made for the change in general 

 conditions during the test. This test, it should be understood, is 

 not for the purpose of obtaining a working correction for the in- 

 strument, but merely to show if the reading indicated by the aneroid 

 is influenced to any degree by temperature. 



Replacement of defective aneroids. — "Wlien aneroids are shown to 

 be defective and giving misleading data, it is important to all con- 

 cerned that they be discarded for better instruments without hesi- 

 tation. 



Many ships' companies have adopted the commendable practice 

 of having their reserve stock of aneroids thoroughly tested at the 



