32 SECTIONAL ADDRESSES. 



being only part of the cause of the diurnal variations. If for a time 

 we assume such increase in the ionisation that the drift currents are 

 sufficient to produce the variations desired, it must further be shown 

 that the circulation of the currents is of the right form to produce the 

 effects. It is obvious that a drift current of constant value always east- 

 ward would produce a permanent effect but not a variation, and any 

 variation in current intensity will tend to produce an accumulation of 

 charge. In the case of a diamagnetic cap, since the intensity is greatest 

 towards midday we may imagine positive charge teiading to accumulate 

 in the p.m. hemisphere and negative in the a.m. hemisphere. But the 

 cap is a good conductor in the direction of the magnetic field, and a 

 current may pass from high levels to low levels and vice versa by travelling 

 in the direction of the field. Below the diamagnetic layer or drift layer 

 where the free paths are long there is a lower layer or layers where the 

 free paths are short and the conductivity is not so anisotropic, and passage 

 from the diamagnetic layer to such layer or layers and vice versa is assumed 

 to take place along the magnetic lines of force in this way. In the northern 

 part of the sunlit hemisphere the current system will, therefore, be from 

 west to east in the diamagnetic layer, then south to north downwards 

 along the magnetic lines of force, then east to west in the lower layer, and 

 finally north to south downwards to the diamagnetic layer. This, of 

 course, is a general rather than a detailed picture. In the southern 

 portion of the sunlit hemisphere the current in the drift layer will again 

 be west to east, then north to south downwards, east to west in the lower 

 conductivity laj'^ers and south to north upwards to the higher layers. 

 Such a type of current system may be imagined to result from drift- currents 

 or the ' dynamo ' theory. With regard to the relative merits of the three 

 theories, an effect of the diamagnetic layer appears certain, but with it is 

 associated the drift current effect which is much larger. The diamagnetic 

 layer effect must therefore be regarded as secondary in importance. The 

 dynamo theory inA'olves motions of the air as well as ionisation, and while 

 on the whole the drift current theory appears to be superior, more informa- 

 tion is needed on the number and distribution of ions and electrons in the 

 upper atmosphere before coming to a final decision. 



The fact that the foci are not on the noon meridian has not been 

 satisfactorily explained. Chapman has suggested a combination of the 

 drift and dynamo hypotheses as a possible ex})lanation, the dynamo action 

 being responsible for the advance of the foci by one hour, but no good 

 reason for such advance is known. It is possible, however, that the drift 

 current theory may alone be a complete solution. If the earth and its 

 atmosphere did not revolve, the ionisation and the drift currents would 

 still exist, and the foci of the currents would then be on the central 

 meridian. With rotation of the earth and its atmosphere, ionisation 

 effects begin at sunrise and soon afterwards a lower conducting layer 

 is formed or strengthened. In the sunlit hemisphere we may assume, 

 therefore, that the ionisation effects are greater on the p.m. side, and 

 the conducti%nty and depth of the layers are also greater on that side. 

 In the northern hemisphere it is possible, therefore, that the effective 

 S. to N. current on the p.m. side is nearer the meridian than the corre- 

 sponding N. to S. current on the a.m. side. Such asymmetry would 



I 



