INSTRUCTIONS TO MARINE METEOROLOGICAL OBSERVERS 29 



(3) The muslin must be thoroughly wetted when a new piece is 

 to be applied to the bulb, and fitted in the manner already explained. 

 The psychrometer should at all times be handled with great care, 

 and when not in use it should always be kept in the case provided , 

 by the manufacturer of the instrument. The fact that the ther- 

 mometer bulbs project below the metal back makes them very sus- 

 ceptible to breakage. Consequently, when the psychrometer is being ; 

 whirled, the utmost caution should be exerted to make sure that the 

 thermometers have plenty of room in which to swing clear of all 1 

 obstacles. 



BUCKET AND INTAKE METHODS OF OBTAINING OCEAN SURFACE 



TEMPERATURES 



Experience has shown that numerous difficulties exist in the mak- 

 ing of accurate observations of the temperature of the sea surface. 

 Both the bucket and intake methods commonly employed on steam- 

 ships have their own peculiar disadvantages and sources of error. 

 It is the temperature of the sea surface that is desired, but the con- 

 denser-intake method gives temperatures at some distance below the 

 surface. This is not objectionable, however, since the Avater_ between 

 the surface and intake depth is almost always so Avell stirred by 

 wave action that its temperature is essentially uniform to that dej)th. 



The bucket method is always a time-consuming operation and is 

 particularly troublesome on large ships. It is not an easy matter 

 to dip out a sizable sample of water with a bucket when the ship is 

 moving rapidly. Often the bucket can be only partially filled, even 

 by the most skillful handling. (Some vertical stiffening of the 

 ordinary canvas bucket and an extra middle ring, to prevent collapse 

 on entering the water, are desirable.) Further difficulty is encount- 

 ered while the sample is being hauled up to the deck — the force of 

 the wind, which is often considerably augmented by the movement 

 of the ship, may cause the bucket to swing about and strike the side 

 of the vessel, thereby spilling a large part of the water. 



In addition to the difficulties of obtaining a satisfactory sample 

 by the bucket method, there exist a number of influences tending 

 to change the original temperature of the sample. These are, no- 

 tably: (1) The bucket itself is not likely to have the same initial 

 temperature as the sea surface; (2) a small quantity of water left 

 from a previous observation and having a temperature different 

 from that of the sea may be in the bucket when the dip is made; 

 (3) the water sample being hauled up is usually cooled by evapora- 

 tion, especially in the case of a canvas bucket; (4) the thermometer 

 inserted is seldom at the same temperature as the water in the 

 bucket; (5) while the thermometer is resting in the bucket further 

 cooling, or perhaps heating, of the water may take place, and (6) 

 sometimes the temperature of the water drawn up has been affected 

 by the discharge through ejection pipes. However, it is possible 

 to avoid these undesirable effects to some extent. How this may be 

 accomplished is described on pages 49 and 50. 



If observations by the condenser-intake method are carefully made, 

 the only source of error of any consequence arises from imperfect 

 installation of the intake thermometer. If the latter is at a point in 



