The 22-year cycle in geomagnetic activity is characterized by high act
ivity during the second half of even-numbered solar cycles and the fir
st half of odd-numbered cycles. We present new evidence for this 22-ye
ar cycle using the aa magnetic index for the years 1844-1994. Over thi
s 150-year interval, the 22-year cycle can be observed through differe
nces between the decay phases of even- and odd-numbered cycles in (1)
average values of a 27-day recurrence index; (2) the results of a chi(
2) ''event'' analysis of 27-day recurrences of both disturbed and quie
t days; and (3) an apparent annual modulation of the 27-day peak in th
e power spectrum of the aa index. Currently, the 22-year variation is
attributed to the Russell-McPherron solar wind - magnetosphere couplin
g mechanism working in conjunction with the Rosenberg-Coleman polarity
effect. Contrary to this viewpoint, we argue that an intrinsic 22-yea
r solar variation (other than polarity reversal), revealed in the syst
ematic low-high alternation of even-odd sunspot maxima within the last
six complete Hale cycles, is the dominant cause of the 22-year cycle
in geomagnetic activity. This sunspot and related. coronal mass ejecti
on variation should lead directly to higher geomagnetic activity durin
g the first-half of odd-numbered solar cycles. Various lines of eviden
ce (including 1-3 above) indicate that 2?;day recurrent wind streams a
re more prominent during the decline of even-numbered solar cycles, co
ntributing to the higher geomagnetic activity observed at those times.
These stronger recurrence patterns may be related to the more rapid e
xpansion of polar coronal holes (faster movement of the coronal stream
er belt to low latitudes) observed following the maxima of recent even
-numbered cycles, The amplitudes of the 22-year sunspot and geomagneti
c activity cycles over the last 150 years are shown to be highly corre
lated. The 22-year pattern of geomagnetic activity appears to be a ref
lection of the solar dynamo coupling of poloidal magnetic fields on th
e decline of one solar cycle to the toroidal fields at the maximum of
the following cycle. It seems likely. that the 22-year variation in su
nspot/solar wind activity plays a role in the observed 22-year modulat
ion of galactic cosmic ray intensity.