GEOPHYSICS: L. A. BAUER 
27 
standing residual effect such as to alter the daily mean values of the 
magnetic elements by an amount 10 to 100 times that caused by the 
regularly-progressing secular variation. The magnitude of the effects 
may at times exceed the average ones described in {a) and (6), depen- 
dent upon peculiar local conditions (ionizations) of the upper atmos- 
pheric layers. Changes in solar radiation may thus furnish sufficient 
cause for the ever-present minor perturbations and elementary waves, 
or pulsations, of the earth's magnetism. 
d. The daily non-cyclic changes in the earth's magnetism, as found 
on magnetically-quiet days by previous investigators, furnish an addi- 
tional check on the foregoing results, their quantities harmonizing com- 
pletely, both as regards sign and magnitude with those given here. It is 
found that on consecutive quiet days the magnetic constant is, on the 
average, larger on the second day than on the first, the increase being 
equal to that which would be caused by an average daily change in the 
solar constant. Moreover, the reason why the magnetic constant, or 
the horizontal intensity, is larger, on the average, on the second quiet 
day, is because, on the average, the solar constant is slightly smaller 
on the second day than on the first. The relation between solar change 
and magnetic change during consecutive quiet days is precisely of the 
same sign and amount as given in {h) . 
e. If the quiet day magnetic effect were to persist throughout the 
year, it would cause a secular variation fully ten times that generally 
observed. However, the quiet days are in the minority, being exceeded 
three times and more by unquiet days, on which the magnetic effect is 
of an opposite or compensating kind to that of the quiet day. Since 
these acyclic effects appear to be associated with solar changes and since 
the latter are not periodic, but more or less sporadic, there is an out- 
standing effect at the end of the year which causes an irregularity in 
the regularly-progressing secular change. Accordingly, there should be 
found some correspondence between annual changes of the solar con- 
stant and annual magnetic changes. This is found to be the case. 
Since the solar-constant changes occur only approximately in accord- 
ance with sun-spot activity and since the magnetic changes are found 
to conform so closely to those in the solar constant, an explanation is 
obtained as to why the irregularities in the magnetic secular change do 
not always synchronize with changes in solar activity as measured by 
the sun-spot numbers nor correspond in magnitude to them. 
