190 BULLAKD AND MASON [CHAP. 10 



wave limit of this gap (Bullard et al., 1962). The existence of the gap and its 

 long-wave limit at about 2000 km is also suggested by a study of some profiles 

 from Project Magnet by Alldredge et al. (1961), though the method of analysis 

 does not allow any detailed conclusions to be drawn. 



4. Reduction of Magnetic Observations 



The magnetic observations may be affected by the magnetism of the ship 

 and by time variations of the earth's magnetic field, as well as by the geology. 

 If the magnetometer is more than two ship's lengths astern of the ship, the 

 observed field wdll require no correction for the magnetization of the ship. This 

 condition is easily realized in practice. Three types of time variations must 

 be considered, the normal daily variation, magnetic storms, and the long-term 

 secular variation. 



The daily variation of the total field on land is usually of the order of 40 y, 

 and a correction for this is desirable. A correction depending on records from 

 the nearest observatory on land may be used, but it is doubtful if this really 

 imj^roves the consistency of the observations. The induction of electric currents 

 in the ocean will tend to reduce the daily variation at sea below that observed 

 on land. The calculations of Chapman and Whitehead (1923) suggest that the 

 reduction will be to about two-thirds but Rikitake (1961, 1962) has pointed out 

 that the effect will be reduced if the mantle is a good conductor of electricity 

 at a depth of a few hundred kilometres. Observations on oceanic islands some- 

 times show a small daily variation; for example, the mean daily variation of 

 the total field at San Miguel in the Azores during the months of May, June and 

 July, 1958, was 30 y, whilst at Coimbra in Portugal, on almost the same latitude, 

 it was 53 y. A few measurements with a towed magnetometer, at a point 

 between the two, suggested that the daily variation in the open ocean was 

 nearer that in the Azores. However, data from oceanic islands must be regarded 

 with caution because of local effects arising from the induced currents flowing 

 in the ocean round them. Observations by R. G. Mason (unpublished) on the 

 island of Oahu show that at points only 20 miles from the Honolulu magnetic 

 observatory the amplitude of the daily variation can differ from that at the 

 observatory by as much as 10%; on averagely disturbed days hour-by-hour 

 differences of more than 20 y are not uncommon. Anomalous effects might also 

 be expected near the boundaries of oceans and continents. Observations by 

 Hill and Mason (1962) suggest that in deep water off the end of the EngHsli 

 Channel the daily variation is greater than it is in western Europe. The theory 

 has been investigated by R. B. Roden (unpublished) \^'ho has shown that such 

 an effect is to be expected, especially on the east side of an ocean. The increase 

 is caused by the same current system as produces the decrease in mid-ocean. 



A daily variation of up to 200 y has been found at sea near the equator 

 (Oguti and Kakinuma, 1959). This high value is related to the equatorial 

 electrojet system of the ionosphere. The effect of the electrojet is at a maximum 

 near the magnetic equator and dies away within a few hundred kilometres 



