3384 HYDROGRAPHIC MANUAL PaGE 242 



Owing to the more rapid rate of change in altitude, the vertical angles measured 

 on east and west celestial bodies are less accurate than those on north and south ones. 

 Observers are likely to have a personal error, so that they always mark a little late or 

 a little early. This seems to be more pronounced for stars at greater altitudes and 

 when the vessel is rolling, and is probably due to the rotary motion customarily given 

 to the sextant in attempting to make certain that the observation is marked when the 

 sextant is in a vertical plane. 



The effect of an error in time or a personal lag in observation is to translate all 

 lines of position east or west of their true positions by an equal amount. Such an 

 error is not eliminated by a series of star sights selected to plot in a symmetrical figure 

 (see 3387A). 



Experience has shown that positions derived from star sights taken by several 

 observers and worked up independently almost invariably differ more in longitude 

 than in latitude, and that the difference between any two observers is likely to be 

 consistently of about the same magnitude and in the same direction. 



C. THE IMPORTANCE OF A DISTINCT HORIZON 



A distinct and clear horizon is probably the most essential of all requirements 

 for obtaining acciu-acy in astronomic observations. Only in very clear and calm 

 weather is the horizon perfectly distinct and sharp. In stormy weather when the 

 height of the seas makes the horizon noticeably irregular, better results may be obtained 

 by taking the observations from as high a station as practicable in order to increase 

 the distance to the horizon. In weather which is not perfectly clear, dark clouds low 

 on the horizon often render it indistinct and at times sizeable errors have been made 

 in observations by mistaking the upper edge of a low narrow band of cloud or fog for 

 the actual horizon. In hazy weather, or when the horizon is indistinct, observations 

 from a low elevation will give better results. 



The brightness of the horizon is a most essential requirement in stellar observa- 

 tions. For this reason it is extremely important that evening stars be taken as early 

 and morning stars as late as practicable, when there is sufficient daylight to illuminate 

 the horizon reasonably well. There is a period of 15 to 40 minutes, varying with the 

 latitude, during morning and evening twilight which is best for stellar observations. 

 In extremely low latitudes rapid observing is necessary to get the required number of 

 observations in the limited time available. The use of predicted altitudes and azi- 

 muths of selected stars will assist greatly in getting the evening observations while 

 the horizon is still sufficiently bright (see 3387B). 



D. OBSERVING PRECAUTIONS 



For best results certain precautions should be taken in observing astronomic 

 sights, which will give additional accuracy. The observer is stationed on a vessel, 

 which is pitching and rolling more or less at all times. This disturbs the observer 

 and continually varies the height of eye. To eliminate the effect of this as much as 

 practicable an observer should station himself near the centerline and amidships. 

 When the vessel is pitching or rolling, each observation of a series should be marked 

 at a moment when the effect is at a minimum and the vessel is on an even keel. 



Astronomic sights should not be observed from a station where the line of sight 

 passes through or near the rigging, or near the masts, smokestack, etc., because the 



