theoretical model may not apply over shorter paths such as 

 those from Criggion to the Netherlands sites. While the 

 phase differences observed in Holland cannot be explained 

 at present, the path lengths, 285 and 281 n.mi. (528 and 

 521 km), are comparable as are the residual discrepancies. * 

 The foregoing is not an adequate sample from which 

 to verify the theoretical model. However, it is at least en- 

 couraging and may constitute an initial understanding of 

 Omega operation near transmitters. It is noteworthy that 

 both the observed and theoretical discrepancies are small 

 at moderate distances from stations. Exclusive of sites 

 where higher modes may be significant, the rms discrepancy 

 of the observed data in this region is 7. 2 cec if no second 

 mode corrections are made. 



A consistent picture of arctic propagation cannot 

 be expected from brief monitoring at only five locations. 

 Arctic However, some of the measurements in the Canary Islands 



propagation and the Azores are of considerable value since the residual 



errors can confidently be said to include the effect of a de- 

 crease in the propagation velocity of the Haiku signal in the 

 vicinity of Hudson's Bay and Labrador-. In particular, the 

 propagation paths may be compared with a recent conduc- 

 tivity map by E. L. Maxwell** from which it is found that 

 about 3000 km of propagation occur over ground with a con- 

 ductivity of one millimho per meter or less, nominally 

 averaging 0,6 millimho per meter. While theoretical esi- 

 mates of the additional phase delay expected for a 10.2-kc/s 

 signal propagated over this path are not directly available. 



*Criggion-Summit phase differences in Portland, U.K. , r 

 fleet propagation over a very short path. 



**Furnished in a personal communication to the author, 

 9 June 1965. 



18 



