sign in sign up
<<
Home , Caloris Planitia, Davies (crater), Hollows (Mercury), Mercury (planet)

VOL. 80, NO. 17 JOURNAL OF GEOPHYSICAL RESEARCH JUNE 10, 1975

Some Cømparisonsof Impact Craters on and the

DONALD E. GAULT,x JOHN E. GUEST,9' JOHN B. MURRAY,2 DANIEL DZURISIN,a AND MICHAEL C. MALINa

Although the general morphologiesof fresh mercurian and lunar craters are remarkably similar, comparisonsof ejectadeposits, interior structures, and changesin morphologywith sizereveal important differencesbetween the two populationsof craters.The differencesare attributableto the differentgravity fieldsin whichthe craterswere formed and have significant implications for theinterpretation of cratering processesand their effectson all planetarybodies.

INTRODUCTION impactvelocities, possibly thermal history, etc., may have con- With additionalevidence from the 10 photography tributed some different effects, but the influenceof these of Mercury[Murray et al., 1974a,b], as Well as the recent variablesshould be of second-orderimportance, at best,i n observationsof [Rumsey et al., 1974],it is now firmly es- comparisonwith the gravitationaleffects. There are, for exam- tablishedthat impact crateringhas been a geologicprocess of ple, compellingarguments [Murray et al., 1974b,1975] that the primary significancein the evolution of all the terrestrial fresh mercuriancraters, which are of interestfor comparative (and undoubtedlyall planetary objects).The heavily purposes,have been formed in silicatesat least grosslysimilar cratered surfacesof Mercury, the moon, and document to those of the moon and, hence, would provide similar the earliest stagesof their planetary histories,and it is clear physicalproperties and responseduring crateringat the scale that a thoroughunderstanding of crateringprocesses and for- of impact events of interest (diameters greater than a few mation is essentialfor gaining further insightinto the early kilometers).Similarly, although impact in excessof history and subsequent development of all the terrestrial 130 km/s are possiblefor Mercury, the averagemercurian im- planets. Although the early cratering record of has been pact velocitiesfor comparativepurposes are probably in the erasedirrevocably and lost by the action of other more active range of 25-30 km/s, within a factor of 2 greaterthan typical geologicprocesses, the recent and relatively young terrestrial lunar values [Wetherill, 1975]. Higher impact velocitieswill impact structureshave been important for an understanding provide increasesin the vapor and melt productsof the mer- and interpretationof lunar (and martian) cratersand cratering curian events,but the effectson the styleof crateringand final processes[e.g., Shoemaker,1962; Baldwin, 1963]. Conversely, crater morphology should be negligible. studies of lunar (martian) craters offer a means for com- It is the purpose of this paper therefore to present first parisonswith terrestrialcraters and crateringphenomena in resultsof somecomparisons of the morphologyof mercurian general [e.g., Hartmann, 1972]. However, suchempirical ap- and lunar cratersin order to explore differencesin cratering proachesto the studyand evaluationof crateringon different style and final landforms caused by what are probably planetary bodieshave limitations;impact crateringis a com- predominatelydifferences in gravitational forcesand induced plex physicalprocess that is dependenton many parameters, stresses.In what follows, the morphology and examplesof and unless the functional dependences are known or mercuriancraters which were studiedin detail are followedby recognizedand relationshipsare established,cratering criteria a brief considerationof scaling(crater size) relationships, com- developedfrom cratersformed in one planetaryenvironment parisonsof quantitativemeasurements of ejectadeposits sur- may not be applicable to similar processesand effectsin a rounding craters, comparisonsof statisticsof interior crater different environment. Moreover, erosion of martian and ter- morphologicalelements, and discussionwith implicationsfor restrial craters by agents foreign to the moon has been so planetary cratering in general. severethat it seemsquestionable whether valid comparisons MORPHOLOGY OF MERCURIAN CRATERS can be made between craters on the three bodies; that is, the lack of fresh, sharp impact structureson earth and Mars The morphologyof mercuriancraters is grosslysimilar to precludes,or at leastobscures, evaluation of any differencesin that of lunar craters;i.e., the morphologicelements (rim facies, satelliticcraters, inner terracedwalls, central peak(s)and inner the cratering processesand crater forms arising from differencesin their planetary environment. mountain rings) that characterizelunar impact cratersare all In marked contrast, however, the mercurian craters from associatedwith craterson Mercury. As with the moon, it can the photography provide comparisonswith lunar be demonstratedthat with increasingage the morphologyof craters under almost ideal 'experimental' conditions.With no cratersbecomes less sharp until most of the crater elementsare appreciableatmosphere having envelopedeither body since barely recognizableand the crater is degradedto a low rim the earlieststages of their history nor any fluvial action superposedwith large numbersof impact craters.The domi- [Murray et al., 1974b,1975], only one prime crateringvariable nant denudation processis apparently, as on the moon, the has been different between the two crater populations, sameprocess which producedthe original crater, erosionand gravitationalacceleration. Differences in physicalproperties, sedimentationby meteoritic bombardment. The freshestcraters away from the have well- developedand extensivebright ray systems,many of them ex- • Ames ResearchCenter, NASA, Moffett Field, California 94035. 2 University of London Observatory, Mill Hill Park, London, tendingfor hundredsof kilometersand a few extendingin ex- England. cessof 1000 km (Figure 1). Some cratersare also surrounded 3 Division of Geological and PlanetarySciences, California Institute by dark halos resemblingthose associatedwith such lunar of Technology, Pasadena, California 91109. craters as Tycho and Aristarchus; these are best seennear the Copyright¸ 1975by the AmericanGeophysical Union. limb underconditions approaching zero phaselighting. Other 2444 GAULT ET AL..' MARINER 10 MISSION 2445

.. %'•-•-"...... •.- }, •. :-'"•:•:•,? :'7'.;.::'•.':'.7 '.': '., '

.•:•::':""....:-•3•.....--.:>•:•"•', • •,'" .:.-....-•.-'•< ,- ...... , •...:.:•,• ': ..>Y:'•.•::•...... -. %' • ", .. C"

:.-...... •'....•:: ?• '...:...... •::•.,•: •.'•.' .'-.•. • '• •'• :--:•:•....:.•f: •.-...... •.•::•".:' .'.•..• .. ':', , :•.• , .• ' ...... , .

.?½:y....::..: :.

..•,..• . ., ,. . r•:••.... ,.• :5' • "'. • -:'.:•.½ , ,:' ...."'::' ,.-•" ':'"•....• ":• : )• •' .•;•:..:.,•? ..•.:..:-•::.•. :..•,..::•..;.:-•::•.•) t. •,. ... • , -7:.•':.:..:•-.,...... •.;...... :• •

,. •.• ., , • '•

...... :.)

.•......

ß . • :•.."•.,.,. •..: ....•.•,-•.-.... t::: "'•;•6::.::....,:•:?:"•• .." •,...... ,.'•'.' •...... • - ...... -:•-:t ;::...:.:...... •...• ..... , :::' '" ::::•'"'• :•..* . '.c': 4"•: -. ::• ..:'.-;: ..-½::...,•.• ...... :., >,.,.. • .. , •, • .: -•, :::::•:•:.....•;•:•...... :--.:-:...... :•:•:. :•..•.:.:: g:•'-.½ :•.•..-...::-,:•..,..-..' t•.--,-•y½ •% ,3•.....:•' ,.-, • ' ,, :'•.-•;:•::.-:... :;:• ..:':.•....-;•::•., - •,.,.. >.•,•...... ,..

;.• ::.:••,:•;• ..:•,•,•1 • ..•:.. .,::•:•.-•

%..':•::...::• •. ½•,.., ,.... ,. •:.'•. ':::- ••-•:•_•: • .? "• :• 2.•:, -•7: ;:: :?•" .:• ....::•'":":'s- ....•'•:•-..•'.•-:•.•''•t'::"•_&•: "-.. .• I -•:...,.-•:?"•:,• •. '..,,:,:• ' ':' '-" •"' .•...:• '7?Z.:f::"::*.'.'* ":'•:-•4'::•:•' :':*'..

.... '::,•¾•'--:';•,...,•.• :•: G-f?.•', '?-•'"' ."'" ': ....'•':'..,':•- •:'•:::/•:::•' :s'•:-.:':-•;•;- •:",-. •'•':• - ' ::•..:2..... :•:.. ""*-'':-• L .•f.• .. . .•--::•, ' . :- ...... •::•:•:.:..-• •f:-•• •-.•• ,•: ,-:•.•' ..-• ::.•::..•::,.:.• .... :,•?.•::•.,:' :}•:>,...:::•.• ß .... •::::•':•:•:•::*':k:•,,.:.: 3-...::..Li[ ,•'•.:: "::..> ...... • .:::½,,:.:•:::•...•,.:.-•.:. ::•;:::.•:.-•::.•::.,• '-'::•:•.....•. .[:••..:- :....• ...... •:::•.... :,•!• ....•: .:,•.:• ,.•,•..•-•..' ,, ., '::"• • • :••:•:• ...... • "-:; • -:...... /• •-• •:• ., • ..... •; • .... •:. .•::..•:,:::-:• -.•,•:• •::•:• _

:.:•::.•4!'::'::•:• ' •..--•;.• :'::•.7 :':•-.': ' :"'•'," .....======.... - •:•-•:•'• "•:"' :k'"•"'•;•: ':'•:" .....-:.::•...... :::- • ;.•,:•-,,...•:•;•:• :,. •..-• ...... :::

...... Fig. la. Preencounterview of with eveningterminator.

Fig. 1. Photomosaicsof Mercuryshowing location of cratersselected for detailedstudy of their morphologyand ejecta deposits. Coordinates and diameters of craters are listed in Table 1. 2446 GAULT ET AL.: MARINER 10 MISSION

.• ß,, ,, .½'j.:r':f.;.½•..'½ :•* ...?'..• ...... •.:.•:::½ .•..: ':"{'""'":;J'-'• -:--?;•...--.-.-:•.•.'.;•:,,,- •. •• "• '24 ":'e' '':"; ';":::'::,,' ;;::"{ZI:•;:;• '...... •-•,•--•-•%?M:! , .--.'½"½:'""• :...... ::?'[.. •%:•:•;•'•":*: ':•:.•-':-•:""-"..:•

': ...... "" -;L'.• •i•'*-"' :x"'Z:,..,.: -•" ...... •.-:,.&•?-':':'L<•f.•..:•::•--• ....: ½• •,• • . ,•-:.•--., :•' •"-:-'-.'-. ..'•'..:::'.;?':z:a"'"'.,•.•....??•..•:,.:,•:? ...:?, %..':::..:':;•;?• ..•-•.• •. :?..-...... ,? c::.•'• .... .-.;½;.•;:..; .:.?:.. ,

. ., ..,.•...'.•;.•':.?:, .•..-::?. ß...... -- •,.:....,,...•: ,•...... ,..•:...... ,.½...... , ß• •-%.

, • x %...... •.?"•..½'• ...... '...... •.½ ,•;:•. . ..'. •'..... ,;,•:";--.•;•..•. •'-:.•....•...... •' ..": .'?•"i•.... •'•'•? ;"',-• .. •, '••:• •-.,.•.• ,.'.•'-•?:.r•. ,•.:••--, ,.Y;.:'"'•-•:4...... •'••.,::,;:•.½• ..... •,•.... ?'..":.. •-.•,;•".:..<• •.- ,??•4..:•',"'... :...-;p:•¾'•;..:.'-----• ...., ..;•.;:"•...... " ?4 '• ½ ,.. •'.';½-.;•x..:½•:•, ..•;.;::::{;V¾'".,?$:'% - ',:-,•:;.•j..•..,•:•.•. '...... ,...;•?:' •._,.•' . ½•:..•.. • , ½•;..• .:.. ,.•,.,•?.•::-.•.: :;,•,•;x,-•. ;.,..::{..,:••..:•a%•.: • :•gj:-" • ,•. %--½-.,.•;;'::,•. .?',.,•.• :• ' .• '..•.. :'.:•;:'; .•-•½• ;.•:.•::.;'• •";•...•!4 ...... •...... :}:•:-' , "•. '%': .... , .%•.• .... :?-'•.... •., ?J•'•..;::• -•"½ •;.;•"-;½•t, "' -- '•. •-: ' •'.'•'•-':•' e.:"i•...... -•."'.. -':'<." •' .: T'., •"• . ½•'•'•R?:."' ...-;½*

•,.'"• ,• ..,.,•,::....<•-::?• ..... ½j:..:;?....•.•,., :½..:...... ?-' .... ,:...- ,•:•;;..½;,...... :j•.: .• .,.•-.,,:..•½.•.,-:, ..• ...... e...... ,x •..... • :-... ::•...... ,...... •...•;'• ...... •...... • .....,?;/:-• •..,- .... .::..: ½. .-*•J • .'•: '. ,.• .. ,-:,.:.½ •½.•.:•..,•.:.,•,.•.•;•?,-:--.- •--'•:7.:....•:.•. •.:..• ;:..-::,•t•'...: .....•,•,-•.•.• • ;':.•;;-•.½*.... ½'..{'•;••.?• ½•;,....• ...,-:•;:•;;' ½ -:'•":•.'.:•."•"'•?,..,%•, '%' .... •"• $ • •'•.'"'•' '"':;:4:•%.J""• :,';J.":'7• ...... :,::•" ;::•":<'•...... ,..•.•:"•:;•:½ -•:B%"• '•.... .'½'-;;.•: '...... •.•---?'• ' ' '½...... ;>?'..... -• :'*?"•:;½ ...... • ?' (...• ,,•%½.•..,."•r :-..•"•½...:'..•:.• '".:,.• ...... x....•.:.•,..,".•,.' .... ?7' ...... ,':;: • •:x,,•,' ' ?..... ': ?•:• ';'t • ß ' .;?•"?:'"½-"'"•:•:'-"•'.•'•¾" .. ß,,.•,.. .,• .•.•.:. • ...... <•.. ..., ß .. •....•...... • -..;.- •..ß •,½ • •'"...... ¾'X.• ::½:?, .•:.?:•..,-.... ;;,..c (.,:..,.;•.,;:.• •.•' ,& •;.....• ...... ,-' •:,..•., ,.....•*• ".•.,• ,..:•%•...... • . •..--•. •:.:::• .,.. -.:-..:.•.. .;½•...-½- ..... • • ,-* ...... •.•ß' ' %.;r:.,,..• ..•,•;:...... • • ,,..• q:.;;•,•'-•'• -:•; ...... --..:. •.•½...•½:...... :.•;.- .::::-..-,.....

'•, •-•:.,.ßß ,...... ß ...... •. .,• :.., • . ...v. ,••' Fig. lb. Postencounterview of planet with morning terminator. GAULT ET AL.: MARINER 10 MISSION 2447 fresh craters appear to have lost their ray systemsand dark secondarycrater fields reveal that the craters are less circular halos, but they remain fresh in all other respects. than normal primary impact craters and that the craters are The morphology of the outer rim units of the mercurian often elongatewith their long axis radial to the primary crater. craters distinguishesthem from their lunar counterparts. At The herringbone pattern and V-shaped features associated earth-basedresolution of the moon, which is comparable to with lunar secondary craters [Guest and Murray, 1971; the available resolution for most of the mercurian craters, the Oberbeckand Morrison, 1973] are not generally found on the rims of lunar craters appear to consistof an inner hummocky mercurian craters probably becauseof resolution limits, but faciesthat gradesout into a radially ridged facies [e.g., Guest, on one of the highestresolution pictures, such features appear 1973]; these facies make up what is termed the continuous to be present (Figure 4), suggestingthat this phenomenon is ejecta blanket and grade outwards into a zone of satellitic not unique to the moon. craters forming crater clusters and chains, interpreted as The interior structure of mercurian craters is dependenton secondaryimpact craters and discontinuousejecta deposits. size, much in the same manner as has been noted before for On the mercurian craters the hummocky rim facies is well- lunar craters [Stuart-Alexander and Howard, 1970; Hartmann developed but grades out over a relatively short distance and Wood, 1971]. The smallercraters are predominatelybowl- through the radially ridged faciesinto a zone characterizedby shapedcavities, but with increasingsize some central peaksare satellitic craters (Figure 2). The areal of craters in this evident, and slump features develop on the walls and may fill outer region is very high and givesthe impressionthat the sur- most of the bottom of the crater (Figure 5). For still larger face is close to saturation [Gault, 1970]. For craters with cratersthe inner slumping forms well-developedterraces that diameters greater than approximately 150 kin, crater chains do not extend across the entire crater floor; the floors tend to are extremely well developed to give long linear grooves be hilly, but nevertheless,flatter and rising from them are (predominately radial to the crater) with crater form, scalloped central peaksor clustersof peaks(Figures 6 and 7). The largest rims; in some casesthe grooves extend across the rim facies craters, which would be called basinsin lunar terminology, are near the crater rim crest (Figure 3), a phenomenonrarely seen characterized by a ringed complex of peaks or a complete on lunar craters. High-resolution pictures of the mercurian mountainous ring concentricwith the crater rim (Figure 8).

,.•:-•::.,:-.9.. . ..

. . :

, ß.. •..? L'.:.:"".. "•...... i:'•.

...... '*:::-,.:?-.';•.• .:...--'..,•:•'5? ...... , . '" ,.. ::? -.""i;;,..:•.:,: ...... :;• }:',:: ....3.."L.:-"•:-•... ",:-.

.,...... ,

.:,

.. : .:.:•;;':'"•:'!*•:ii:.:•½•.,?.... '! ß' ..'•":"'7& •' .. ;:."?::.. ...•. ;' ,•. !::':!?--;:, ,...... '-::. %-...::...;•'•"•-;-:;;.:;'>' ;7 ,'• .::. --.'-.....•::.,, ......

.....•:.:.....•

'•"•.:•;:.; -,.•:::...:.

-:½ :..¾.•':....:•:•%•... ,:::• -..:.,--.:,%½,:•:•' .:.. . ß:: :,:•.:;:..- .. "'-',4-•:•? .:..:' .. {:::(:•'• ...... Fig. 2. Crater 16 (98-km diameter)illustrates the narrow hummockyrim facies,radial ridges,and surroundingexten- sive field of secondarycraters. The well-developedinterior terracesand central peaks are typical for mercurian cratersin this sizerange. Note that the smallercraters in the foreground(--•25-km diameter) also are terraced.Flight Data Subsystem (FDS) frame 80. 2448 GAULT ET AL.: MARINER 10 MISSION

tablishinga scalingrelationship. Starting from the reservoirof kineticenergy KE representedby the motionof theimpacting bodyrelative tothe surface ofthe planet, energy isexpended for five processes:heating, En; comminution,Ec; plastic andviscous deformations, E;,v;removal (ejection) ofmaterial from the crater, Ee; and seismic waves, Es [see Gault and Heitowit, 1963; Braslau, 1970]. Although the shock compres- sions produced by impact are adiabatic processes,the as- sociatedincreases in internal energyof the compressedmasses are not accomplishedisentropically (entropy constant), and in- ternal energy is trapped irreversibly as heat after release of stressesto ambient conditions. This energy is manifested in melting and/or vaporizationof both the impactingand the im- pacted material. The fraction of KE converted to heat is a functionof the materialsand the impact Vt and can be expressedhere as En = k•f(Vt) (1) with k• a constantandf(Vt) a function of impact velocity.It is worth noting that En representsa fraction of the KE that is lost to the cratering processbecause it does not contributeeither directly or indirectly to removing material from the site of the '•'' .•:•i!!i .•'., : . impact during formation of the crater. , •'-,:.,...-- •...•-.:.• .• '.•'...:•. ::-•.-/-: ...... Comminution energy, the creation of free surfaces, is directly proportional to the volume of material crushedand ß:..•.:•? : . .• •., . . ..:..•:. .... :•,,;:.•..•:•...: broken up during the crateringevent and is a function of the •ig. 3. Orthographic proj•ctio• of •DS ffam• 166 size distribution of the comminuted .Taking the volume 140-km diameter crater 12 a•d its surrounding zo• of s•co•dar• of the crater proportional to the cube of its diameter D (i.e., craters. Th• •arro• •idth of th• rim faci•s• th• prominent subradial the shape of the excavationcrater is independentof size; see m•rcuria• craters. sectionon depth-diameterrelationships), then it can be shown that the comminution energy can be written as Ec = ko.Da-ø'• (2) The sizesof mercurian craters in this progressionof interior structuresare systematicallyless than for lunar craters;quan- with k: a constantand wherea and •i are positivevalued terms titative values will be presentedlater. used to describe the size distribution of the crushed material of

CRATER SCALING the . Similarly, the energy expendedfor plastic and/or viscousdeformation as the embryonic crater develops There is firm theoretical and experimental evidencethat to its final excavationsize will be proportionalalso to the vol- gravitational accelerationis an important parameteraffecting ume of the crater. By introducing another constantks we ar- the size of craters formed by either explosivesor impact rive at [Viktorov and Stepenov,1960; Chabai, 1965; Johnsonet al., 1969; D. E. Gault and J. A. Wedekind, manuscript in prep- E•,v = kaD• (3) aration, 1975]. Thus an important difference that must be In contrast,the minimum energynecessary to remove(eject) anticipated to exist betweenmercurian and lunar craters,for the excavatedmaterial from the crater is proportional to the which the Mariner 10 imagerycannot provide a basisfor com- product of the crater volume, someaverage height the volume parisons,consists of the effectsof on the dimensionsof must be raisedto depositthe ejectaaround the crater, and the their respectiveexcavation craters. Here, by definition,excava- tion crater is taken to mean the crater attained as the direct gravitational acceleration. Becausethe height can be taken proportional to the diameter, the minimum ejection energy result of processesrelated to stress(shock) waves which set becomes material in motion to form the crater, in contrast to the final crater, which is the excavation crater modified by postcrater- Ee = k•gD• (4) ing processesunrelated to the passageof stresswaves. Because expected differencesin the excavation craters on the two with a fourth constantk• and the gravitational accelerationg. bodies are significant to subsequent discussions,a brief Finally, the shock waves produced by the impact attenuate qualitative discussionof crater scaling criteria is given in the with increasing distance from the point of impact and following paragraphs. For a more rigorous treatment of ultimately decayinto a complexpattern of seismicwaves con- variablesand problems of crater scaling(although the discus- taining a fraction of the original KE. The functional sion is concerned specifically with craters formed by ex- dependenceis unknown, but for presentpurposes it can be plosives),the reader is referred to Chabai [1965]. written as A simple approach to demonstratethe variations and ap- Es = k,KE (5) plicable rangesof various scalingrelationships is to consider the partition of energyfor impact events;that is, we will con- Although this seismicenergy is a very small fraction of KE (k, sider how much and in what manner energyis expendeddur- • 10-• to 10-•), a Caloris event on Mercury or an Imbrium ing the formation of an excavation crater as a means of es- event on the moon (10• ergs?)could producea seismicwave GAULT ET AL,: MARINER 10 MISSION 2449

. :.•.•,, • ::=:' . .• .-,,.• •..,..•= , ,,,=:.• ..... :• .,,'.•'•,•;...•.'-.• .. ,• .... . :.,:=-ß :•. .• ':';:•, ....

.( •'..,t•:•;•, '.• ...... •.,. - • ?-.' ', •:..:.•' ....?. ?, •.." :t:?, .":•' .,.,.' ....':'•: ':,;-' '.,. ,•"•"'•-'• .;•.•' ,.?' v,...... t.','".'•. , "."-" ...••'•'. .,• ...... •"..•,''.?•':' "'.'¾•,Z.• •'.,'?•.•-•' •'?"•'•'.. •'...? '"•'•,.:, • ....• "'•,."•.'•--'.',•' ,.?" •:.,.'.... ., ...... • . :"..•-.

,.•-- ..•. ' .... ß ...."' .-., "'•, -'" •'•.• .....,•'..•' .' "," '• .•-•..... •,•' •" ß...... 2 ' '., ',•'"•- , • •.. ",'",'...•.....•' ' •. ß•'. •...•.:' .... '• ',,-":.. .:•"• '• '•'-:,,.•..• .v .. •, ., .• • "',• •",..• .•., ..• ':• ,•.• , •.?:-' ,' . •"'-.' '":•.•-'.,: "":;' , ':..?'•. ß'• ,-...•...•-•'""•"•'•. ... . ',, .- '.':,•.,..:•,•' .•?•'•" "h•.•:' .,d•,:•,'.•.• •,,

,• ..... ,,•:•. "?•• '••, ...' .....,• . .• ...,•..:•. .-,..,::'•..,...... :,:•...... ,• •: ...... ,... . . • •....•..•

' ,•-,•'•"•-:'..•:-'•-'•:•,' ,•'•.• ...... •'•-•: -•,':•'.•.• .•.,.•..', '•.•,- ..,•:•'•",",•.'• •."•,':•',•' • ' •;..•:•'Z:,•, ßß .-• •':"• ':•'--'"•'-.•.• "?,•*:' ß ',.'•:,"'•., m• •:•-'...•"-•'"' •-:•" .....•,• •'•'•,•"""?• "•.'•: '• • •.•' ,,.,....•. ' •,•' "-.•. ....,..c•,. • ,:'",'•, ,'•%'?•, •.?,• "'"'• .... •' ,.,,,'••-•;• •,•'..•. •.?•. •.• '•,',.•..•'• •'"...•,• •:•'?"•'•"...... •,'• ...... , ' '•.... "-'•"•?-.": ' •?'?'•.• •. '"' ,•'"•-•" •-•.'...... ' ,•,C?'.•?, •.•,,.,.• "''• •+ ...... •r"'•,•'•'.'•.,'

'.•, ß .•:. ' .•",'" •,,- '• - ' ,•.'•.. •..:•.• . * .r .,'•.... • • .•.'-• -•'•.•.-..-"..• ' •: '- ß .. : •. -•. ,..•.•..• ,• ... ";., '•"'"• '?'•;•?'•?"• •?"••'•'•" •"ß '•'?"•'?•'•'•"'•"'•'•--'•.•'•'•v•"• '--•',,•?-'--'.... '•"•.•-•:'" ,,.'•• ...... '•:• .-•...... •. ',,.•?•..,.•...• ,, ,•.,.-,.:.•,•...... •.. .. •.•-.-•..•'•,,.)•,:,.,...•.-%•' •'...... •.•%--' •.• ß ...... ?.. '?,% .....•,. -,. .:•'..,.?;.,'•' ,...... , ...... ß . ,? -•,.::.• .....•?• ,. •.•. ,, Fig. 4. On• o• th• highestr•solution pictures(fidd of view, 104 • 62 km) showinga w•ak h•rringbon• pattern and V- shapedf•atur•s in associationwith 2- to 3-km diameter craters(lower right corner), apparentlyproduced by clustersof s•condary•cta from an unknown sourc• ot th• southwest(north up). FDS • •1. train orders o• magnitud• greater than th• intensity o• any th• hilly and lin•at•d t•rrain d•scrib•d by Track and Gue• known t•rr•strial •arthquak•s. Som• speculationson th• [197•] that is antipodalto th• Calorisbasin. •ff•cts o• s•ismic•ff•cts o• suchmagnitudes ar• discussedby Takingth• sumo• th• fiv• •n•rgy•xp•nditur•s and •quating Schultzand Gaul• [197•], in particularth• p?sibl•r•lation to th•mto th• KE r•sultsin

Fig. 5. Orthographicprojection of FDS frame 27459 showingthe morphologicprogression from small bowl-shaped cratersto largercraters with incipientterracing and centralpeaks. The freshsharp crater (centered) is the 20-km diameter crater 7 with Mercury'ssurface reference crater Hun Kal [Daviesand Batson, 1975]on its southernflank. 2450 GAULT ET AL.: MARINER 10 MISSION

for velocities in excessof about 15 km/s (D. E. Gault, un- publishedresults, 1969). For this reasonit is convenientto col- lect terms involving just KE, F(KE) = KE- k,f(Vt)- ksKE and rewrite the energy balance as F(KE) • k:D8-• + ksD• + k4gD4 For craters in rock, k: will have a value of the order of 109[see Gault, 1973], while k4 will be lessthan 104[Gault and Heitowit, 1963] but greater than 10-• consideringthe minimum energy necessaryto excavatea crater; on the basisof explosivecrater- ing data, k• • 10L Thus k: >> k• >> k4, so that for small crateringevents on planetarybodies involving modest amounts of KE the two terms on the right become insignificant, and the diameter D is related to KE by D o: F(KE)•/(a-"•) (6) while for very large impactsthe terms involving D a becomein- significant and D o: g-X/4F(KE)m (7) Fig. 6. Crater 17 (74-km diameter) just north of the displaysinterior terracingand centralpeaks rising up from a with a transitional range between the two extremes of hilly floor. The continuousejecta deposits and secondarycrater field are well-.ieftned.Orthographic projection, FDS frame 79. D o: F(KE),/a (8)

KE = kff(Vt) + k:Da-"6 + kaDa + k, gD' + ksKE The last equation is the sameas the well-known cube root scal- ing relationshipof Lampson[1946], and (7) expressesa fourth If the KE is held constant as the impact velocity is increased, root dependence,the so-called gravity scaling relationship the fraction of KE converted to heat remains relatively con- [Chabai, 1965]. The preciserange of diameters(i.e., KE) over stant (approximately 30-35% for silicatematerials of interest) which each equation should be applicable is poorly defined,

::, ,'"'•"?:::{•:•",, •;:•;•.":•'''""C •'"*,.•"' ':7:'7' ...... )•,??, ..•:..:.:'•:.:;•:•.•, '-•,•,•,..:•. , •;••::•.

"' -' ;:' ';' -'"'-;•' ': ' ? a•,:;:'";;• /'-?: :*'-5.

Fig. 7. Detailed view of a cluster of central peaks on the flat floor of crater I (120-km diameter). FDS frame 27461. GAULT ET AL.: MARINER 10 MISSION 2451

Fig. 8. Orthographic projection of craters 14 (128-km diameter) and 15 (195-km diameter) that have interior rings of mountains and ejecta depositswhich are scarredby deep secondarycrater chain grooves. FDS frame 150. although (6) is appropriate for craters with lessthan a 10-m Wetherill [1975], the 4-fold increasein KE would compensate diameter formed in rock in the terrestrial gravitational en- for gravityeffects and causemercurian excavation craters to be vironment [Gault, 1973]. For presentpurposes, however, (7) is larger in the ratio 1.15. appropriate to the size of cratersconsidered in the following EJECTA DEPOSITS discussion(>1 kin), and it indicates that for large craters formed with equal KE on Mercury (g = 370 cm/½) and the All craters studied in this and the following sections are moon (g = 16•' cm/½) the mercurian craters of excavation relatively fresh, sharp forms that are included as rayed would be smaller in a ratio of 0.81 due to the difference in craters, dark halo craters, or young craters with well- gravitationalfields. However, for equal massesimpacting with developedejecta and secondaryfields on the geologicmap by velocitiesthat typically differ by a factor of 2 as suggestedby Trask and Guest [1975], comparable to craters of Copernican 2452 GAULT ET AL.: MARINER 10 MISSION and Eratosthenianage on the moon or class 1 cratersin the TABLE 1. Location and Size of Craters Used for Continuous systemof Arthur et al. [1963].Detailed studies of the ejecta Ejecta Blanket Measurements depositsof themercurian craters have been limited to 27struc- Latitude, Longitude, Rim Diameter, tureswhich were selected to giveas broad a coverageof crater Crater deg deg km sizesand morphologictypes as possibleconsistent with ac- ceptableresolution of the imageryavailable from the firsten- 1 -15.95 16.03 120 2 21.02 18.27 240 counterwith Mercuryby Mariner !0. Resultsfrom the second- 3 20.14 21.51 120 encounterphotography were not processedin timefor this 4 10.74 21.24 88 study,but theywill providea wealthof additionalcraters for 5 0.41 23.74 90 analysisand a better statisticalbase for data. 6 6.31 14.44 59 7 0.08 20.15 20 Figure 1 showsthe location of the 27 craters: l l craters 8* - 10.94 31.48 63 viewed as Mariner l0 approachedMercury and 16 craters 9 -16.72 28.25 14 viewed after encounter. Table 1 lists the size and latitude and 10 -39.60 31.52 106 longitudeof thesecraters as a referencefor otherworkers. It 11 -44.96 18.42 27 shouldbe notedthat althoughthe westernrim of crater26, the 12 65.60 170.33 140 13 61.15 138.81 70 largestselected for study,has not emergedfrom the morning 14 59.41 127.56 128 terminatorin the photomosaicshown in Figure1, subsequent 15 54.51 136.38 195 photographytaken 14-24hours later revealedits full diameter 16 59.53 178.26 98 and the probablepresence of a centralring similarto other 17 50.20 178.54 74 18 37.86 126.17 54 mercurian structuresof comparablesize. 19 31.40 176.10 77 Continuousejecta deposits. The radial extentof the hum- 20 25.65 180.32 60 mockyfacies between the craterrims and the outerlimits of 21 9.60 133.54 90 theradially ridged facies, the continuous ejecta deposits, was 22 3.55 138.59 120 mappedon acetatesheets overlaying rectified (or originalun- 23 -1.83 146.47 126 24 9.77 169.45 48 foreshortened)pictures of the 27 craterslisted in Table 1. 25 9.12 174.11 49 Three individuals (D.E.G., J.E.G., and M.C.M.) mapped 26 8.19 189.64 250 each crater independentlyand ascertainedsubjectively an 27 - 31.07 168.91 70 averagevalue for the radial extent of the continuousejecta deposits.The averagesof the three determinationsare Location of craters is accordingto the Mariner 10 surfaceco- ordinate system[Davies and Batson,1975]. presentedin Figure9 wherethey are comparedwith similar * Crater [Murray et al., 1974b]. data for lunar craters derived(by D.E.G. and J.E.G.) from lunar orbiter and Apollo imagery. The comparisonreveals that for a givenrim diameterthe depositson Mercurywould have a rangethat is a factor0.44 less than the same material would attain on the moon, a value radial extentsof the lunar depositsare systematicallylarger than thosefor Mercury.For a simplelinear relationship significantly smaller than the observedvalue •Rcm/Rct = 0.65. The two values,however, are not directly comparable,since betweenradial extent of the continuousdeposits Rc and the the formercorresponds to total ballisticrange Re for material rim diameterDr, leastsquares fits to the two setsof data give ejectedfrom withinan excavationcrater of diameterDe, while (Re/Dr),,,= 0.44 (9) the latter is the distance from the rim of craters with diameters Dr to the limit of the continuousdeposits; Re is greaterthan and R•, and Dr may be considerablygreater than De, dependingon (lO) the amount of postcrateringmodifications that may haveoc- (Rc/Dr)t= 0.68 -- 1.5 )< 10-aD,. curred. Moreover, the excavationcraters De for the two cases for the mercurian and lunar cases,respectively. The expres- will be differentbecause of the originalassumption of identical sionsare appropriateonly for craterssmaller than approx- conditions(i.e., KE) exceptfor gravity fields. imately 300 km. The comparisonimproves if the hypotheticalconditions are Taking the ratio Rc,/R•t for equalsize craters, (9) and (10) yieldmercurian ejecta deposits that areapproximately 0.65 the width of lunar deposits.This reducedwidth of the mercurian depositsis qualitativelyconsistent with the relativevalues of the gravitationalaccelerations for the two bodies,but in order •l / CLASSICRATERS ßMERCURY to makea morequantiative comparison, consider first two im- pactevents that occurunder identical conditions (i.e., material •-tu L '•' x EARTH properties,impacting mass, impact velocity, etc.) with the ex- •-I=_.5 .[•••••• ß ....,_•....Q.. '•'•- ceptionthat the gravityfields are different. Peak shock stresses and wavegeometry will be identicalfor the two events,so that x ejectavelocities, angles of ejection,fragment size distribution, 0 5• •00• •50• 200I 250 etc., should also be identical.For this hypotheticalcondition RIMDIAMETER, km the ballisticrange Re• of ejectareaching the limits of the con- Fig. 9. Average radial extents of continuousejecta deposits tinuousdeposits of themaximum extent considered in Figure9 around freshclass I cratersfor Mercury, the moon, and earth. Lines (<100 km) is adequatelydescribed by the simplerange equa- through mercurianand lunar data points are least squaresfits to measurements;earth data pointsare basedon Ries basin,Germany, tion for a flat surface.On this basis the ballistic range'is in- Meteor Crater, Arizona, and laboratoryhypervelocity (6-7 kin/s) im- verselyproportional to the gravitationacceleration Re• ocl/g, pactcraters formed in noncohesivesand. See text for discussionof line and ejecta comprisingthe outer limit of the continuous through terrestrial data points. GAULT ET AL.'. MARINER 10 MISSION 2453 now changedby increasingthe size (KE) of the impacting curian craters and then counted in concentric annuli around body which impacts the surface with the higher value of the primary structures. Four lunar craters were similarly gravitationalacceleration. The relativewave geometrywill re- mapped by using lunar orbiter and Apollo photographs. main the same and scaleup directly with the size of the im- Figure 10 presentsa map for mercuriancrater 12. Becauseof pactingbody; the new excavationdiameter De will increaseac- the difficulty in distinguishingbetween secondary craters and cordingto (7), and the absoluteposition of the effectivepoint small primary impact craters near the resolutionlimit of the of ejection of material from within the enlarged excavation pictures,all craterswith diametersgreater than an arbitrary crater will be movedradially outward. When the ratio Re/De is lower limit are includedin the counts(0.014 of the primary rim now formed by using • as the exponent for the scaling diameter).However, for the lunar cratersAristarchus, Coper- relationshipsin (6), (7), and (8), nicus, and Harpalus [Murray, 1972], secondarycraters could be distinguishedwith considerableconfidence on the basisof the lunar orbiter and Apollo imagery. Thus while the lunar and the mercurian-lunarratio Rem/Retfor equal valuesof De countsare probably most representativeof the actual second- becomes 0.54 in better agreement with the observations ary crater populations, the counts for Tsiolkovsky and the Rem/Ret.This larger value, however,is still not directly com- mercurian craters may include some superposedprimary parable to the observationsbecause for the generalconditions craters.In all cases,many of the small cratersin the pictures representedin Figure 9, postcrateringmodifications have oc- are difficult to differentiatefrom and depressionsof curred (i.e., Dr > De, as will be shown later in the section on other origins, particularly in the rough hummocky terrains interior structures),and there is evidencethat the increasein near the primary craters;counts on all sevencraters were done

diameter for a given initial value of De is not the samefor Mer- by one individual (J.B.M.),, so that resultspresented herein are cury and the moon. subjectivelyconsistent. Perhapsa potentially more meaningfulcomparison can be Only the secondarycrater counts for crater 12 are complete. made by writing the rangeRe in termsof Re. By introducing Due to its proximity to the eveningterminator, crater 1 is the incrementalincrease in diameterAD due to postcratering poorly illuminatedon its easternside, and countsfor only the modifications and a constant 0 < n < 0.5 which defines the westernside could be accomplished.Only countsto the eastof positionwithin the excavationcrater at which the ejectaeffec- crater 19 have been completedby January 1975. For all three tively leaves the crater, mercurian craters, other large primary craters were in relativelyclose proximity. Some of thesecraters had fieldsof Re = Rc + nDe+ AD (12) secondariesclose to the fieldunder study; where these areas Forn = 0 theejecta isconsidered toleave the excavation crater could be identified, they have been excludedfrom the counts at its rim; for n = 0.5 the ejecta would originate from the presentedhere. center of the crater. The ratio R•m/R•t can then be obtained in Results of the secondarycrater counts(Figures 11 and 12) the form confirm the generalimpression gained from visual inspection R•= D•,,,(R•/ D• -- n -- A D/D•),,, and comparisonsof mercurian and lunar craters; the areal densityof the mercuriansecondary crater population is greatly Rot D,t(R,/D,-- n -- AD/D•)• (13) enhanced over those for the lunar craters. However, the Becausevalues for AD are unknown, consider only those generaltrends of the secondarypopulations both on Mercury smaller craters(D < 10 km) for which AD = 0, so that De = Dr and on the moon are very similar. With increasingdistance and De,,, - Der, and recast the ratio into R• (kg•-• -- n)• = R t (kg•--1 where k would be the constantof proportionality in (11). BecauseD•m = D•t, this approximation effectivelyassumes from (7) that KE• = (gm/gt)KEt;for impactingbodies of the samemass this would correspondto conditionswhere the im- pact velocityagainst Mercury would be a factor of 1.5 greater than on the moon. For the limiting conditionn = 0, (14) givesthe sameresult as (11) (i.e., R•/R•t = 0.54), and for n • 0 the value nm for o .:-•./::' . ø%Oo 0'• Mercury must be lessthan the value nt for the moon if the ratio R•m/R•t is to be greaterthan 0.54. Without someknowledge of the value of the proportionaiRyconstant k, however,the ap- propriate valuesfor n to achieveor approacha value of the ratio of 0.65 cannot be obtained. On the other hand, if the ,•o • • o.,• .... o.o. • fraction of the total ejecta mass that is containedwithin the continuous deposits can be estimated, values for n can be derived,and possiblysome evaluation of the proportionality constantk can be obtained--a potentiallyimportant approach I I I for future studiesto determinescaling laws of large impact o õoloo events. SCALE (APPROX.) Secondaryimpact craters. Secondaryimpact craters were Fig. 10. The field of secondarycraters surrounding crater 12 as mappedonto clear acetatesheets overlying photographically mapped from an enlargedrectified image of FDS frame 166 shown in enlargedand computer-rectifiedpictures of threeprimary mer- Figure 3. 2454 GAULT ET AL.'. MARINER 10 MISSION

30 , ! ! ! 30 Tsiolkovsky are comparable for diametersbetween 110 and

I 140 km(?) if the Tsiolkovskyhas experiencedchanges in its • MERCURY I I jCRATER NO.•2 diameter proportional to those of crater 19. •o 20 For the two pairs of primary craters(Figure 11) the radial

o variations and numerical values for the ratio of mercurian to I I MOON lunar secondarycraters are virtually identical. The ratios are • •o I •-• TSlOLKOVSKh •0 greatest(> 10) nearestto the crater rims and are clearlyrelated to the difference in the extents of the continuousejecta deposits;the mercuriandeposits are narrower and evolveinto the secondarycrater fieldscloser to the rim than in the caseof 2O0 400 o 2oo 400 DISTANCEFROM CENTER OF CRATER,km lunar craters. The ratios decreaserapidly, however, with in- creasingdistance from the rim, as the crater countsare based uponthe populationsof both secondaryfields. Values of about , -30 , •, , , ] 1 are attained at a distance of approximately two crater % ,oo MERCURY (77km) J diameters from the rims of the excavation craters and CRATERNO.,91•:20 thereafterremain essentiallyconstant or continueto decrease z o 3 slowly for another crater diameter, the radial limit of the data. MO, The greatly enhancedareal densityof mercuriansecondaries o•0 relative to lunar populationsis explicable, again, in terms of the differencesin the gravitationalconstants of the two bodies. o Qualitatively, the reducedballistic range on Mercury acts to o 200 a: 0 I00 200 DISTANCEFROM CENTER OF CRATER,km reduce the dispersionbetween individual ejecta fragments producing secondariesand servesto concentratethem into Fig. 11. Radial variationsin the areal densityof secondarycraters and the ratios of mercurianto lunar secondarycraters for Mercury smallerareas circumscribing the point of the impact.The net craters 12 and 19 and for lunar craters'Copernicusand Tsiolkovsky. effect leadsto a greaternumber of secondariesbeing formed per unit area. In this connectionthe long linear crater form grooves(Figure 3) are a result of this reduceddispersion for from the crater rims the areal densityof secondarycraters in- the large secondaryejecta ; in a lower gravitationalfield creasesrapidly to a maximumvalue and then decreases more the secondarycraters would not overlap each other but, in- slowlyover a distanceof severalcrater diametersto valuesless stead, would produce a chain of discretecraters. thanthose at the rim. The increase in densitywith distance It shouldbe cautioned,however, that for both comparisons from the rim is most rapid for the mercurian craters, but the shown in Figure 11 the mercurian rim diametersD• are less position of maximumdensity, both for Mercury and for the than their lunar counterparts.Because of the arbitrarycutoff moon, occursoutside limits previouslydetermined for the rim in the countsof secondarycraters at a size0.014D•, numerous faciescomprising the continuousejecta deposits, a resultcon- smaller secondaries have been included in the mercurian sistent with the distinction made between an inner rim facies counts that have been excludedfrom the comparable lunar and an outer zone of satellitic craters. data. This exclusion will serve to increase the mercurian secon- The maximum values of areal density for the secondary dary crater frequencyrelative to the moon by an unknown crater populationsincrease with decreasingdiameter of the amount.For thisreason it is instructiveto considera simple primary crater,but this is an apparentresult caused by the ar- model for estimatingthe effect of concentratingmercurian bitrary limitation invokedto limit the secondarycrater counts to craterslarger than 0.014D•. Thus the cutoff for Aristarchus and Tsiolkovsky occurs for craters of 0.5-km and 2.5-km .•40 , , , diameter, respectively,and would correspondto a factor % MERCURY between l0 and 102decrease in the relative crater population z for the latter. Comparisonof the secondarycrater populations betweenMercury and the moon thereforemust be made for o 20 craters of approximatelythe same size. Toward this end the secondarycrater countsfor craters12 (140 km) and 19(77 km) z are compared in Figure ll with approximate dimensional o analoguesin the lunar environment,Tsiolkovsky (180 km) and 0 200 400 Copernicus (90 km). Although the craters do not have the DISTANCEFROM CENTER OF CRATER,km

i i i i i i same rim diameters, all four craters have experienced •160 postcrateringenlargement of their excavationdiameters due to slumpingof the interior wallsas indicatedby terracing.The in- _ MOON • MOON cremental changes are rather uncertain, particularly for N /ARIsTARCHUS I I• HARPALUs

Tsiolkovsky,whose interior structure has been modified by en- -- 80 dogenetic processes[Guest and Murray, 1969], but it is es- timated that the excavation craters 12 and 19 have been l••L•(4økm) • I (4'km) enlargedby about 30 and 13 km, respectively,and Shoemaker [1962] suggeststhat there has been at leasta 25-km increasein 0 ioo o ioo the diameter of Copernicusdue to slumping.Crater 19 and DISTANCEFROM CENTER OF CRATER,km Copernicustherefore provide a comparisonfor craters with Fig. 12. Radial variations in areal density of secondarycraters for transient diameters De of about 65 km, and crater 12 and Mercury crater I and lunar craters Aristarchusand Harpalus. GAULT ET AL.: MARINER 10 MISSION 2455 secondariesinto smaller circumscribingareas around the i i , , ! ]/I I ' , 120•- // EARTH "1 primary crater. [ / (NO )l For this application the ballistic range equation for a L /// •-""""'" .... sphericalbody is requiredbecause the maximumrange of the data (400 km) exceedsthe limits for a flat surfaceapproxima- tion. Figure 13 presentsthe ratio of mercurianto lunar bal- 20' •" listic range as a function of the ejection velocity for two representativevalues of the angleof ejection,which includes the valuessuggested by laboratory impact experiments[Gault et al., 1963;Gault et al., 1968;Stbffier et al., 1975].The ratio is !,o relativelyconstant (0.44) and insensitiveto ejectionangle up to velocitiesof about 0.5 km/s (mercurianrange • 60 km; lunar .... i .... i .... i .... i , 0 .5 1.0 1.5 2.0 • 150 km) but decreasessignificantly for the highervelocities, EJECTIONVELOCITY, km/sec especiallyfor the more shallowangle(s) of ejection.Velocities Fig. 14. Calculated areal density of secondary craters for of about 1.2 km/s are necessaryfor the 400-km range on Mer- atmosphere-freebodies relative to the moon, basedon valuesof bal- cury. listic range given in Figure 13. See text and (16). Now, if we considertwo cratersof approximatelythe same size, as discussedpreviously, then the population of the flagmentalmaterial producingsecondary craters and the con- simplifiedmodel used to correct for the differencesin the ditions for ejection from the primary craterswill be similar, gravitationalacceleration, the measuredand calculatedvalues and the family of mercuriansecondary craters will impact the of the enhancementof mercuriansecondary crater population surfacein concentricrings closerto their source,as shownby are in reasonableagreement only at the outer limits of the con- Figure13. The effectivereduction in the referencearea Ar for tinuousdeposits. Beyond about two craterdiameters from the the population at a given range Re will be inverselypropor- rims, however, the simple model grosslyoverestimates the tional to the radial distance from the crater center, or in terms ratio even though the basic data are biased toward some of De and n, enhancementof the mercurian secondarycrater population. Ar (z [Re + (0.5 - n)De]-•' (15) The oversimplificationof the model would be first suspectfor the disagreement,but one cannotdiscount the possibilitythat Appropriate values for n will approach 0.5 becausethe a fundamentalassumption behind the model is not valid, in material ejectedat the highestvelocities originates closest to particular , that the materialproperties and fragmentsize dis- the impactingbody at the centerof the crater.For thisreason, tributions from lunar and mercurian craters are similar. whenRe is largerthan De by a factor of more than 2 or 3, (15) Further studiesand comparisonsare required to substantiate may be approximatedas A• 0: Re-•', and the areal densityof thesenumerical results, but the qualitativeeffects of increased the mercuriansecondary craters compared to densitiesfor the gravitycompressing and restrictingmercurian secondaries into moon will be increasedby the ratio a narrowerring aroundthe point of impactseem adequately confirmed.

At=_ R• 2 (16) INTERIOR STRUCTURE

Numerical values for the areal densityratio, basedon Figure As the basisfor quantitativecomparisons of the interior 13, are presentedin Figure 14 and mirror the rangeratios. The structures of mercurian and lunar craters, 130 mercurian areal densityof mercuriansecondary craters is approximately craterswere selectedfor depth-diameterdeterminations and 5 timesgreater than that for the moon for the shorterranges the occurrenceof terracingand centralpeak(s). This groupof correspondingto ejectionvelocities less than about0.5 km/s, cratersincludes many of thoselisted in Table2 and is essential- but the ratio increasesrapidly for velocitiesin excessof about ly randomlydistributed with respectto planetarylatitude, but 1 km/s. For the extremerange of 400 km for the comparison they clusterwithin approximately20 ø longitudeof the morn- shown in Figure 10 the secondarycrater ratio would be ingand evening terminators in orderto obtainshadows which between7 and 9, dependingon the angleof ejectionfrom the are usefulfor the measurements.All craters correspondto crater. Within the constraints of the crater count data and the class1 featuresin the systemof Arthur et al. [1963]based on the subjectivejudgment of thedegradationa! state indicated by the crispnessof the rim and interior structuresand the freshnessof the continuous ejecta deposits and/o r secondary craterfields. Depth'diameter calculations from shadow length analysisare basedon the best spacecrafttrajectory and camera-pointinginformation available in January1975; limitedby thequality of thesupporting data, the reliability of the measurements is estimated to be better than 10%. Depth-diameterrelationships. Results from the mercurian depth-diameter measurementsare presented in Figure 15, where they are compared with Pike's [1974a] most recent o .s •.o •.s 2.0 EJECTIONVELOCITY, km/sec photogrammetricdeterminations for lunar craters.Although the available resolution of the Mariner l0 photography Fig. 13. Calculated ballistic range for ejecta on atmosphere-free restricts the data for the mercurian results to craters larger bodies relative to the range on the moon as a function of ejection velocity and angle of ejection0 measuredwith respectto the local than I km, it is neverthelessapparent that the trend of tl•e horizontal. depth-diameter relationship for craters on Mercury is 2456 GAULT ET AL.: MARINER 10 MISSION

I0: ' ' '• .... I , . , i .... I ' ' ' I .... I ' ' ondUnits, 370/162 = 2,3)and with the suggestion [e.g,, : CLASSI CRATERS • ' I",• Shoemaker, 1962; Quaide et al., 1965; Gault et al., 1968] that postcrateringslumping of crater walls is induced by gravitationalforce s andleads to interiorterracing and perhaps central peaks. Slumping would serveto decreasedepths of cratersrelative to their diameters,and mostimportantly, : • ßMERCURY slumpingwould be expectedto firstoccur for somecritical size of crater for which the induced lithostatic stresseswith increas- I0-1 o• [3MOON (PIKE, 1974•) • o oø ing diameterwould exceedthe local material strengthsaround ß o •:•DD0TM and under the transientexcavation crater, thus precipitating slope or base type failures [Varnes, 1958; Mackin, 1969]. If 10-2 . ,., .... I , , . • .... I , , ß • .... I , , ! , suchan interpretation is valid, then terracesand central peaks I0-I I I0 I02 I0:3 on Mercury also shouldoccur at smallerdiameters than on the RIM DIAMETER,km moon. Accordingly,the frequencyof their interior structures Fig. 15. Comparisonof the depth-diameterrelationship for fresh mercurian and lunar craters. as a function of crater diameter is consideredin the following section. Terraceand centralpeak frequency. Figure 16 and Tables2 remarkably similar to that for the moon. There is a and 3 presentthe frequencyof terraced and central-peaked pronounced kink in the mercurian data set at diameters mercuriancraters for the samesample of 130 cratersused for between 5 and 10 km in a manner similar to that for the moon depth-diameter determinations. Similar data reported by at diametersaround 15 km. For diametersless than, say, 7-8 and Sanchez[1973] for lunar cratersare shownfor com- km the mercurian craters appear to have a depth-diameter parison, and it is evident that both terracesand central peaks relationship very nearly identical to that of the moon; for do occurin smaller craterson Mercury than on the moon. For diameters greater than 7-8 km the mercurian craters are Mercury, 80% of the 10- to 20-km diameter size classcraters systematicallyshallower than their lunar counterparts,but are terraced, and effectivelyall craterslarger than 20 km are they do follow a trend almost parallel to that for the moon. terraced;only 2 out of 71 craters(3%) larger than 20 km did If we follow the procedureof Pike [1974a] and divide the not have terraces.In contrast, only 12% of the 10- to 20-km mercurian data set into two groups, a least squaresfit for sizeclass of lunar cratersare terraced,and completeterracing diametersless than 7 km (the diameterthat yieldsminimum (equivalentto the mercurian3%) is not attained until crater residuals)gives diametersare greaterthan 40 km, twice the sizethat is required d = 0.15Drx'ø* (17) on Mercury. Comparable results also are found for central peaks. For and for diametersgreater than 7 km Mercury, 80% of the 10- to 20-krn size cratersexhibit central d = 0.93Drø'27 (18) peaks,while only 5 out of 71 (7%) with diameterslarger than 20 km do not have peaks. On the moon, the Smith and whered and Dr are, respectively,the depthand rim diameterin Sanchezresults give 26% peakedin the 10- to 20-km aizeclass, kilometers. For lunar craters,Pike [1974a] gives for craters and complete'peakism' equivalentto completeterracing does less than about 15 km, in the same notation, not occur until craters are between 30- and 40-km diameter, d = 0.196Drx'øxø (19) again almost twice the size of their lunar counterparts. Moreover, if allowances are made between the two data sets and for diametersgreater than about 15 km for the differencesin (1) the resolutionbetween Mariner 10 im- ageryand the lunar orbiter and Apollo photographyavailable d = 1.044Drø'aø• (20) to Smith and Sanchez,and (2) their interpretationand defini- The strong similarity between the two suites of data is tion of central peaks,the contrastbetween Mercury and the evidencedby the degreeof agreementin the leastsquares solu- moon in the frequencyof central peaks is even greater (see tions. Over the range of the mercurian measurements,(17) Table 3). yields values 15-20% lower than (19), and (18) givesvalues An additional comparison of the interior structures is 25-30% lower than (20). Direct comparisons,however, should affordedby the resultsshown in Figure 17, where the progres- be made with caution.Pike [1974a]points out that photo- sion with crater size from the occurrenceof simple central grammetric results yield more accurate results than peak to concentricringed complexesis given for Mercury. In shadowtechniques and in generalgive slightlygreater depths. TERRACES CENTRAL PEAK(S) Differencesbetween the two methodsare mostpronounced for I00 the smallercraters below the kink in the data. Thus it appears possible,if not probable,that the differencesbetween (17) and (19) are causedeither totally or in part by techniquedifferences in depth determinations,while the differencesexpressed between(18) and (20) are real and representthe effect(s)of .• so differencesin their crateringenvironments. In supportof this interpretation,independent of the precisenumerical fits to the depth-diameter data, is the observationthat the kink in the o 40 80 0 40 80 mercuriandepth-diameter relationship occurs over a rangeof RIM DIAMETER,km crater diametersthat is approximatelya factor of 2 lessthan Fig. 16. Comparisonbetween Mercury and the moon for the fre- for lunar craters. This is consistent with the ratio of their quencyof occurrenceof terracesand central peaks as a function of respectivegravitational accelerations (in centimetergram sec- crater rim diameter. Lunar data from Smith and Sanchez [1973]. GAULT ET AL.: MARINER 10 MISSION 2457

TABLE 2. Comparisonof Terrace Frequencyfor Mercurian and Lunar Craters

, Mercury Moon*

Rim Diameter, Total Number Percent Percent Number Total km Number Terraced Terraced Terraced Terraced Number

0-10 43 3 7 0 0 6 10-20 25 20 80 12 4 34 20-30 19 18 95 33 7 21 30-40 12 11 92 79 15 19 40-50 7 7 100 87 7 8 50-60 11 11 100 100 7 7 60-120• 22 22 100 100 20 20 * From Smith and Sanchez [1973]. •' 140 km for the moon. general the mercurian craters smaller than about 100-km expendedas the transientexcavation craters collapse after they diameter are characterizedby a single peak or a complex of attain their maximum dimensions. As discussedearlier, crater multiple peaks. Beginningat diametersof about 100 km, dimensions are a function of energy, but the functional however,ringed structuresof peaks,concentric with the main relationshipis not linear. On the other hand, gravitational crater rim, also occur; the limited data sample available in- potentialenergy available by collapseof a crater shouldscale dicatesthat the centralpeaks disappear between 130- and 180- directlyand linearly with the depthof the transientexcavation km diameterand only cratershaving an inner ring of peaksoc- cavity. If this interpretation is correct, one would expect cur for diametersgreater than 180 km. This progressionis centralpeak heightson Mercury to be approximatelytwice as strikinglysimilar to that for the moon [Stuart-Alexanderand high as lunar peaksfor the samesize craters. Preliminary Mer- Howard, 1970;Hartmann, 1972]. Betweenabout 150- and 300- cury data, however, are almost identical with those of the km diameter they point out that craters have either central moon, but Wood's [1973] limited data for heights of central peaksor rings, but above300 km, all structuresare ringed, uplifts (peaks?)of terrestrial craters support the argument. a size that again is approximately twice its mercurian Wood[1973, equations 5 and 6] gives(with h the heightand all morphologic equivalence.It might be noted that based on dimensions in kilometers) their observations, Stuart-Alexander and Howard defined h = 0.026Dr -- 0.26 (21) lunar basins to have diameters greater than 300 km, and perhaps a comparable criterion of 150 km would be ap- h = 0.13Dr -- 0.23 (22) propriate for the definition of basinson Mercury. for lunar and terrestrialcraters, respectively. The changein the The resultspresented in Figures15, 16, and 17, taken collec- heightswith diameteris a factor of 5 greaterfor earth than for tively, indicate that the factor of 2.3 difference in the the moon (0.13/0.026 = 5). With considerationof the limita- gravitationalenvironments of Mercury and the moon is mir- tions in the terrestrial data the factor of 5 is in remarkably rored by the approximatefactor of 2 differencebetween the good agreementwith the actual ratio of 6 for the gravitational two bodiesfor the onsetof the changein their depth-diameter accelerations for the two bodies. relationshipsand the occurrenceof terraces,central peaks, and Although we cannot excludethe possibilitythat what ap- ringedbasins. Moreover, the role of gravity is alsosuggested in pears to be remarkably consistenteffects of gravitational ac- the recentresults reported by Wood[ 1973],who found that the celerationmay be, instead, the sum of multiple effectsand heights of central peaks are directly proportional to crater events, the suggestionby Shoemaker[1962], Quaide et al. diameter. Although Wood concludesthat the linear propor- [1965], Gault et al. [1968], and many others that the tionality implies that peak height is a function of crater- phenomena under considerationare the end product of forming energy,we suggestthat the linear proportionalityis processesdriven by gravitationalforces seems more than ade- insteadthe result of the gravitationalpotential energy that is quately confirmed.

TABLE 3. Comparisonof Central Peak Frequencyfor Mercurian and Lunar Craters

Mercury Moon*

Rim Number Number Diameter, Total With Percent Percent With Total km Number Peak(s) Peaked Peaked Peak(s) Number

0-10 43 0 0 0 0 5 10-20 25 15 80 26(3)$ 8(1) 31 20-30 19 16 84 53 (11) 10(2) 19 30-40 12 10 83 79(74) 15(14) 19 40-50 7 7 100 87 7 8 50-60 11 11 100 100 7 7 60-120t} 22 22 100 100 20 20

* From Smith and Sanchez[1973]. •' Estimatedfor photographicresolution and interpretation comparable to Mariner 10imagery (see text). õ 140 km for the moon. 2458 GAULT ET AL.'. MARINER 10 MISSION

i I i I [ I I [Shoemaker,1963] and the Ries basin [Hiittner et al., 1969], MERCURY 200 the third datum being based on small (30-cm diameter) O CRATERSWITH INNER RING hypervelocity impact craters formed in sand at an ambient ß CRATERS WITH INNER RING AND CENTRAL PEAK(S) x CRATERS WITH CENTRAL PEAK(S) pressure of less than 1 torr (D. E. Gault, unpublished data, 1974). The three data points fall well below the values for Mercury. Extrapolation of the mercurian least squaresfit to terrestrial gravity, in proportion to the moon- Mercury ratio, yieldsthe dashedline in good agreementwith the data points.It is to be notedthat Figure 13 predictsa I I I I I smallervalue (0.16) than the observedvalues (•0.2) in Figure o ioo 20o 300 40o RIM DIAMETER,km 9 in the samemanner that the Figure 13 underestimates(0.44) the observedvalue for Mercury (0.54) in Figure 9. It shouldbe Fig. 17. Changesin the interior structureof mercurian cratersand their inner ring diameter as a function of the outer (rim) diameter. anticipated,moreover, that the secondarycrater population should be greatly enhancedover the crater populationsof Mercury and the moon, especiallyat the limit of the con- CONCLUDING REMARKS tinuous deposits(Figure 14). Thesecomparisons between mereuriah and lunar impact The significanceof the mercurian-lunarcomparisons is not craters, taken collectively,clearly indicate important dif- restrictedto the interpretation of craters on other planetary ferencesin the two populations of craters. The differences bodies. Trask and Guest [1975] point out that the most are attributableto differencesin the gravity fields,but whether widespreadsurface unit on Mercury, mappedas an intercrater or not they are the direct and soleeffects of the gravitational unit, is a flat or gently rolling terrain betweenlarge environmentscannot be determined.The degreeof correlation craters and basins.This unit is heavily cratered with small that is obtained basedon assumingonly gravitativeeffects is craters and is analogousto small areas on the Moon in pre- more than adequate,however, to assurethat gravitationalac- Imbrian plains. Trask and Guest postulatethat the mercurian celeration has been the primary agent for the observed intercrater plains are an ancient primordial surface whose differences.Imagery for fresh craters formed in different lunar analogue has been largely obliterated becauseof the gravity fieldswould be ideal for further studies,and although moon's greater number of subsequentlarge cratersand basins the other terrestrialbodies are unsuitedfor this purpose,pos- [Murray et al., 1974b].An additional reason,perhaps of more sibly future photographyof the satellitesof the major planets importance,is that the greatlyreduced ballistic range of ejecta will provide additional observationaldata. on Mercury relative to the moon would reduce the effec- Notwithstandingsome reservations, it is interestingto con- tivenessof largeevents in obliteratingand/or modifyingexten- sider further the depth-diameterdata as applied to terrestrial sive areas on the mercurian surface. craters. The kinks or changes in the depth-diameter It should be noted that the observed reduction in the bal- relationshipsoccur both on Mercury and on the moon for listic transport of ejecta on Mercury relative to the moon has diametersslightly less than the beginningof the rangeof sizes important implications for two recent analyses which over which terracingand centralpeaks become abundant. For developedmodels of the radial thicknessdistributions of ejecta Mercury the kink is between 5 and 10 kin; only 7% of the aroundlunar craters,with particularemphasis on and applica- craters lessthan 10 km are terraced, but in contrast 80% of the tion to the large basinsOrientale and Imbrium [McGetchinet 10- to 20-kin size classare terracedand peaked.In the caseof al., 1973; Pike, 1974b].The models depend critically on an the moon with its kink at about 15 kin, 12% of the 10- to 20- evaluationof (1) the thicknessof the depositson the rim of the km sizeclass a•:e terraced, and it is onlyfor craterslarger than excagationcrater and (2) the functionalrelationship used to 30-kin diameterthat terracesand centralpeaks become com- describethe changein thicknesswith distancefrom the rim. mon at the 80% level. Thus it appears that for crater sizes Although Pike correctsseveral important shortcomingsin the belowthe changein the depth-diameterrelationship the crater M cGetchinet al. analysis,both modelingstudies are basedon geometryis determinedby the excavationprocesses and depth- data from terrestrialcraters, and empiricalrepresentations for diameterratios are effectivelyindependent of gravity,consis- the rim thicknessand radial decayare derivedwith no correc- tent with laboratory impact cratering results [Gault and tions for the differencein gravitationalenvironments between Wedekind,1975]. For cratersizes above the changethe excava- the earth and the moon. Figure 9, 13, and 14 clearly indicate tion geometryis modifiedby postcrateringcollapse of the rim that lunar rim depositsshould be significantlythinner than ter- structurewith a consequentincrease in diameterand decrease restrial deposits for any given size excavation crater. in depth. Extrapolation of this interpretationto terrestrial Moreover, it is not certain that the functional expressionfor cratersformed in a gravitationalfield about 3 timesgreater the radial changein thicknesswith distancefrom the rim is in- than on Mercury leads to the expectationthat a similar kink dependentof gravity. The subjectis beyondthe scopeof this occursfor diameters1-3 km on earth and is accompaniedby discussion,but it is cautioned that the difference between lunar terracingand central peaks. This expectationis in general and terrestrialrim depositscannot be as great as indicatedby agreementwith known impact structuresand impliesthat the Figure 14, becausethe assumptionsbehind (16) are not valid relativelylow relief of the largerterrestrial craters is the result for the conditions Re << De being consideredhere. It is, of gravity-inducedpostcratering processes in additionto long- perhaps,sufficient to note here only that the expressionsfor termed erosional effects. ejecta thicknessespresented by McGetchin et al. [1973] and Some further support for the greater effectsof gravity on Pike [1974b]are poorly basedand requiremajor revisionsto earth is shown in Figure 9, where three data points for the accountfor gravitationaldifferences between the earth and the radial extentof the continuousejecta deposits are compared moon. with the mereuriah and lunar determinations. The terrestrial The obliterating effect of ejecta from large basin-forming data are estimatesfrom maps of Meteor Crater, Arizona events at great distancesfrom the point of impact has par- GAULT ET AL.: MARINER 10 MISSION 2459

ticularsignificance for the moonand is emphasizedby the ing on the lunar surface:Criteria and implications,Radio Sci., 5(2), measurementsof the continuousdeposits shown in Figure 9. 273-291, 1970. Although these data do not extend up to dimensionsof the Gault, D. E., Displacedmass, depth, diameter,and effectsof oblique trajectoriesfor impact craters formed in densecrystalline rocks, order of the Orientale and Imbrium events, the results Moon, 6, 32-44, 1973. neverthelessindicate that their continuousdeposits would cer- Gault, D. E., and E. D. Heitowit, The partitioningof energyfor im- tainly be no greater than 0.25Dr. If Dr is definedby the outer pact craters formed in rock, in Proceedings of the Sixth ring for thesebasins, then the radial extentsof the continuous HypervelocityImpact Symposium,vol. 2, pp. 419-456, Firestone Rubber Company, Cleveland, Ohio, 1963. depositswould be no greaterthan 200 or 300 km, respectively, Gault, D. E., E. M. Shoemaker,and H. J. Moore, Spray ejectedfrom from the arcsdescribed by the Cordilleraand Apenninemoun- the lunar surfaceby impact, NASA Tech. Note D-1767, tains. If an inner ring is usedas a referencevalue for Dr, the 1963. radial extents would be contained within these same arcs. Thus Gault, D. E., W. L. Quaide, and V. R. Oberbeck, Impact cratering the margins of the continuous depositsfor these basins are mechanics,in ShockMetamorphism of Natural Materials, edited by B. M. French and N.M. Short, pp. 87-99, Mono, Baltimore, Md., restrictedto distancesno greater than about 650 km from the 1968. center of the Orientale basin and about 900 km for Imbrium; Guest, J. E., Stratigraphyof ejecta from the lunar crater Aristarchus, secondarycrater fields must extend beyond thesedistances, a Geol. Soc. Amer. Bull., 84, 2873-2894, 1973. conclusionthat is consistentwith and supportsthe argument Guest, J. E., and J. B. Murray, and orgin of Tsiolkovsky crater, lunar farside, Planet. Space Sci., 17, 121-141, 1969. of Oberbecket al. [1974] that secondarycratering and the as- Guest,J. E., and J. B. Murray, A largescale surface pattern associated sociatedmixing of ejecta with local material occur at suchdis- with the ejectablanket and rays of Copernicus,Moon, 3, 326-336, tancesrather than the depositionof a discretelayer of ejecta 1971. from the basin. On this basis the radial lineations described as Hartmann, W. K., Interplanet variations in scale of crater lmbrium sculptureat distancesof more than 1000 km from the morphology Earth, Mars, moon, Icarus, 17, 707-713, 1972. Hartmann, W. K., and C. A. Wood, Moon-origin and evolution of outer ring cannot be part of a 'continuous' deposit. By multi-ringed basins, Moon, 3(1), 3-78, 1971. analogy,the continuousdeposits surrounding Caloris Planitia Hi•ttner, R., H. -Koler, and W. Treibs, Ammerkungen zur must have an even more restricted radial extent, and the GeologischenObusichtskarte (Beilage l), Geol. Bavarica,61, greater portion of the circumscribingterrain representsthe 451-454, 1969. Johnson, S. W., J. A. Smith, E.G. Franklin, L. K. Moraski, and D. J. secondarycrater field from Caloris. Becauseof almost iden- Teal, Gravity and atmosphericpressure effects on crater formation tical gravitational fields, it seems reasonable that Caloris in sand, J. Geophys.Res., 74(20), 4838-4850, 1969. Planitiaprovides a remarkableanalogu½ for the great basin Lampson,C. W., Explosionsin earth,in Effectsof Impactand Explo- Hellas on Mars before degradation to its current eroded state. sion, vol. 1, part 2, chap. 3, p. 110, Office of ScientificResearch and Indeed, the fresh mercurian craters should be models for all Development, Washington, D.C., 1946. Mackin, J. H., Origin of lunar maria, Geol. Soc. Amer. Bull., 80, martian impact craters;the absenceof freshstructures and the 735-748, 1969. degraded morphology of craters on Mars constitute mute McGetchin, T. R., M. Settle, and J. W. Head, Radial thickness varia- evidencefor the effectivenessof eolian erosionand deposition tion in ejecta: Implications for lunar basin deposits, and indicate the inadequacyof martian craters for assessing Earth Planet. Sci. Lett., 20, 226-236, 1973. Murray, J. B., The geology and geomorphologyof the Aristarchus morphologic characteristicsof cratering processes. plateau region of the moon, Ph.D. thesis, Univ. of London, London, 1972. Acknowledgments. We are indebted to many individuals, but Murray, B.C., M. J. S. Belton, G. E, , M. E. Davies, D. E. special acknowledgmentmust be made to the membersof the Image Gault, B. Hapke, B. O'Leary, R. G. Strom, V. R. Soumi, and N. ProcessingLab of the Jet Propulsion Laboratory, in particular J. Trask, Mariner l0 pictures of Mercury: First results, , Mosher, J. Soha, and J. Lorre, for the specialprocessing and rectifica- 184(1307), 459-461, 1974a. tion of images required for this study, to our colleagueson the Murray, B.C., M. J. S. Belton, G. E. Danielson, M. E. Davies, D. E. Mariner l0 TV Imaging Team, B. Hapke, B. Murray, R. Strom, and Gault, B. Hapke, B. O'Leary, R. G. Strom, V. R. Soumi, and N. N. Trask, for valuable discussionsand criticism, to M. Davies for his Trask, Mercury's surface:Preliminary descriptionand interpreta- expeditiouscooperation in providingthe planetarycoordinates neces- tion from Mariner l0 pictures,Science, 185(4146), 169-179, 1974b. sary for picture rectification,and to E. Danielson and K. Klausen for Murray, B.C., R. Strom, N.J. Trask, and D. E. Gault, Surfacehistory their sustainedsupport of thesestudies. Acknowledgment is also due of Mercury: Implications for terrestrial planets,J. Geophys.Res., P. Schultz and P. Cassen for helpful discussions,R. Hayes for data 80, this issue, 1975. reduction, R. J. Pike for providing tabulations of his depth-diameter Oberbeck,V. R., and R. A. Morrison, The lunar herringbonepattern, data, and R. B. Baldwin, W. K. Hartmann, and C. R. Chapman for NASA Spec. Publ. 330, section32, pp. 15-29, 1973. their careful reviews of the text. Two of us (D.D. and M.C.M.) were Oberbeck, V. R., R. H. Morrison, F. HSrz, W. L. Quaide, and D. E. supportedunder NASA grant NGR 05-002-117 from the Planetology Gault, Smoothplains and continuousdeposits of cratersand basins, ProgramOffice. Th• workof J.E.G.was supported by the U.K. Geochim. Cosmochim. Acta., 1, 111-136, 1974. Natural Environment Research Council. Contribution 2581, Division Pike, R. J., Depth/diameter relations of fresh lunar craters:Revision of Geological and Planetary , California Institute of from spacecraftdata, Geophys.Res. Lett., •(7), 291-294, 1974a. Technology,Pasadena, California 91125. Pike, R. J., Ejecta from large craterson the moon: Commentson the geometric model of McGetchin et al., Earth Planet. Sci. Lett., 23, REFERENCES 265-274, 1974b. Quaide, W. L., D. E. Gault, and R. A. Schmidt, Gravitation effectson Arthur, D. W. G., A. P. Agnierary, R. A. Howath, C. A. Wood, and lunar impact structures,Ann. N.Y. Acad. Sci., 123, 563-572, 1965. C. R. Chapman, The system of lunar craters, Commun. Lunar Rumsey, H. C., G. A. Morris, R. R. , and R. M. Goldstein, A Planet. Lab., 2, 71-78, 1963. radar brightnessand altitude image of a portion of Venus, Icarus, Baldwin, R. B., The Measure of the Moon, University of Chicago 23, 1-7, 1974. Press,Chicago, Ill., 1963. Schultz, P. H., and D. E. Gault, Seismiceffects from major basin for- Braslau, D., Partitioning of energy in hypervelocityimpact against mation, Moon, 12, 159-177, 1975. loose sand targets, J. Geophys.Res., 75(20), 3987-3999, 1970. Shoemaker, E. M., Interpretation of lunar craters, in Physicsand Chabai,A. J., On scalingdimensions of cratersproduced by buriedex- of the Moon, edited by Z. Kopal, pp. 283-359, plosives,J. Geophys.Res., 70(20), 5075-5098, 1965. Academic, New York, 1962. Davies, M. E., and R. M. Batson, Surface coordinates and car- Shoemaker,E. M., Impact mechanicsat Meteor Crater, Arizona, in tography of Mercury, J. Geophys.Res,, 80, this issue, 1975. The SolarSystem, vol. 4, editedby G. Kuiper, pp. 301-336,Univer- Gault, D. E., Saturationand equilibriumconditions for impact crater- sity of Chicago Press, Chicago, Ill., 1963. 2460 GAULT ET AL.'. MARINER 10 MISSION

Smith,E. I., andA. G. Sanchez,Fresh lunar craters: Morphology asa Viktorov, V. V., and R. D. Stepenov,Modeling of the action of an ex- function of diameter, a possiblecriterion for crater origin, Mod. plosive with concentrated charges in uniform (hOmogeneous) Geol., 4, 51-59, 1973. ground,Inzh. $b., 28, 87-96, 1960.(English translation, Sandia StiSffler,D., D. E. Gault, J. Wedekind, and G. Polkowski,Experimen- Corporation, Albuquerque, N. Mex.) tal hypervelocityimpact into quartz sand: Distribution and shock Wetherill, G. W., Pre-marecratering and early solarsystem history, in metamorphismof ejecta, J. Geoœhys.Res., 80, in press, 1975. Proceedingsof the Soviet-AmericanConference on Cosmochemistry Stuart-Alexander, D. E., and K. A. Howard, Lunar maria and circular of theMoon and Planets,NASA, Washington,D.C., in press,1975. basinstoA review, Icarus, •2(3), 440-456, 1970. Wood, C. A., Moon: Central peak heightsand crater origins,Icarus, Trask, N.J., and J. E. GUest, Preliminary geologic terrain map of 20, 503-506, 1973. Mercury, J. Geophys.Res., 80, this issue,1975. Varnes,D.J., Landslidetypes and processes,Landslides and Engineer- (ReceivedFebruary 7, 1975; ing PractiCe,Publ. 544, pp. 20-47, Nat. Acad. of Sci., Washington, revised 3, 1975; D.C., 1958. acceptedMarch 3, 1975.)