http://www.skyandtelescope.com.au 51and drastically slowing its eruption.
Additionally,themagmastartstocool,
and the dike sidewalls begin to contract,
becausethemagma’spressureisnotas
high as earlier. All of this slowing of
movementgivestimeforgasbubbles
within the magma to coalesce into
singlevoidsthatgrowaswideasthe
dike.Thisstopsthecontinuouseruption
of magma, breaking it up into individual
pulses,eacheruptingadiscretelayerof
pyroclastic material. Small, abundant
pyroclasticconesform,muchlikethose
foundintheMarius Hills.
Thedikecontinuestocloseduring
phase4,restrictingtheeruptionof
pyroclastic particles, but lava flows are
still pushed out, often rich withvesicles
—hollowsthatoncecontainedgas
bubbles. Some of these flows are short,
building up low domes around the
vent,suchasthewell-knownonesnear
HortensiusandMiliciuscraters. In one
case, a small circular lava pond was
builtupandtheescapeofgasinlateral
feeding dikes caused collapse pits along
rilles,andacalderacollapsedatthe
summit—this,ofcourse,isHyginus.
Some dikes intrude into brecciated
spaces beneath impact craters and
spread laterally. In craters smaller than
about40kmacross,thecentreofthe
lava pile tends to thicken (a laccolith),
pushingupanddeformingthefloor
of the crater. In larger craters, the
inflowing lava forms asill—aflatslabof
lava—elevatingtheentirefloorbutwith
less central deformation. An example of
a smaller floor-fractured crater isVitello
in southern Mare Humorum;Humboldt
near the eastern limb is larger and has
aflat,shallowfloor.Incasessuchas
Plato,Archimedes,andHercules,
rising lava finds an easy path to the
surface, submerging the craters’ floors
andcentralpeakswithsmoothponds
of lava.
Earlier I mentioned that the period
in lunar history in which a dike forms
was critical for its future. Evidence
suggeststhatduringitsfirstbillion
years, heat left over from the accretion
oftheMoon,inadditiontoheat
generatedfromthedecayofradioactive
minerals,puttheearlyMoonintoan
extensional stress pattern, stretching
the lunar crust. This made it easier
forcrackstopropagatefromdiapirs
to the surface. Additionally, the warm
Moon could partially melt portions of
themantlemorereadily,generating
voluminousamountsofmagma.This
theoretical conclusion is consistent
withthefactthatmostlunarmare
lavas were erupted between 3.9 and
3.1billionyearsago,andmostsinuous
rilles formed before 3 billion years ago.
AstheearlyMooncooled,its
interior contracted slightly, creating
a compressional environment that
inhibitedmantlefracturinganddike
formation. Small amounts of lava did
reachthesurfacebetween2to3billion
years ago, and even tinier amountsSLeft: Hadley Rille was formed when large volumes of lava flowed for a week or more, eroding
a sinuous channel in the underlying strata. Right: Some phase 4 eruptions produce dikes that
spread laterally beneath impact craters, forcing their surfaces upward and fracturing the floors
like that of Vitello crater located in the southwestern lunar quadrant.
SThe Marius Hills formed when the eruption
rate of a dike’s magma slowed sufficiently to
result in the building of discrete pyroclastic
cones like the ones seen here.erupted until about a billion years
ago, testifying that despite magma
source regions becoming smaller and
deeper, lithosphere thickening, and a
more compressive mantle structure,
a few dikes did reach the surface in
special circumstances. The paucity
or even complete absence of younger
volcanic features makes it doubtful that
transient lunar phenomena or irregular
mare patches are associated with young
volcanism. Perhaps we will be surprised
someday, but the Moon seems to be
volcanically dead. For an observer
spotting traces of ancient volcanism,
however, the moment is alive with awe.CHARLES WOOD is sorry to have
missed seeing the surges of lava flows
spreading across Mare Imbrium. For
more on lunar dikes see the Wilson and
Head paper at https://is.gd/lunardikes.