ion　of　the　villain；　which　had　its　historical　root　in　slavery。　Some　burdens　again　lay　on　the　land；　and　not　on　the　person。　And　finally；　manorial　exactions　could　grow　from　the　political　sway　conferred　by　feudal　lordship。　It　may　be　difficult　to　distinguish　in　the　concrete　between　these　several　relations；　and　the　constant　tendency　in　practice　must　have　been　undoubtedly　directed　towards　mixing　up　the　separate　threads　of　subjection。　Still；　a　general　survey　of　manorial　rights　has　undoubtedly　to　start　from　these　fundamental　distinctions。　　　　　There　has　been　some　debate　on　the　question　whether　the　lord　could　sell　his　villains。　It　has　been　urged　that　we　have　no　traces　of　such　transactions　during　the　feudal　period；　and　that　therefore　personal　serfdom　did　not　exist　even　in　law。（46*）　It　can　be　pointed　out；　on　the　other　side；　that　deeds　of　sale　conveying　villains　apart　from　their　tenements；　although　rare；　actually　exist。　The　usual　form　of　enfranchisement　was　a　deed　of　sale；　and　it　cannot　be　argued　that　this　treatment　of　manumission　is　a　mere　relic　of　former　times；　because　both　the　Frank　and　the　Saxon　manumissions　of　the　preceding　period　assume　a　different　shape；　they　are　not　effected　by　sale。　The　existing　evidence　entitles　one　to　maintain　that　a　villain　could　be　lawfully　sold；　with　all　his　family；　his　sequela；　but　that　in　practice　such　transactions　were　uncommon。（47*）　The　fact　is　a　most　important　one　in　itself。　The　whole　aspect　of　society　and　of　its　work　would　have　been　different　if　the　workman　had　been　a　saleable　commodity　passing　easily　from　hand　to　hand。　Nothing　of　the　kind　is　to　be　noticed　in　the　medieval　system。　There　is　no　slave　market；　and　no　slave　trade；　nothing　to　be　compared　with　what　took　place　in　the　slave　states　of　North　America；　or　even　to　the　restricted　traffic　in　Russia　before　the　emancipation。　The　reason　is　a　curious　one；　and　forcibly　suggested　by　a　comparison　between　the　cases　when　such　trade　comes　into　being；　and　those　when　it　does　not。　The　essential　condition　for　commercial　transfer　is　a　protected　market；　and　such　a　market　existed　more　or　less　in　every　case　when　men　could　be　bought　and　sold。　An　organised　state　of　some　kind；　however　slightly　built；　is　necessary　as　a　shelter　for　such　transfer。　The　feudal　system　proved　more　deficient　in　this　respect　than　very　raw　forms　of　early　society；　which　make　up　for　deficiencies　in　State　protection　by　the　facilities　of　acquiring　slaves　and　punishing　them。　The　landowner　had　enough　political　independence　to　prevent　the　State　from　exercising　an　efficient　control　over　the　dependent　population；　and　for　this　very　reason　he　had　to　rely　on　his　own　force　and　influence　to　keep　those　dependents　under　his　sway。　Personal　dependence　was　locally　limited；　and　not　politically　general；　if　one　may　use　the　expression。　It　was　easy　for　the　villain　to　step　out　of　the　precincts　of　bondage；　it　was　all　but　impossible　for　the　lord　to　treat　his　man　as　a　transferable　chattel。　The　whole　relation　got　to　be　regulated　more　by　internal　conditions　than　by　external　pressure；　by　a　customary　modus　vivendi；　and　not　by　commercial　and　state…protected　competition。　This　explains　why　in　some　cases　political　progress　meant　a　temporary　change　for　the　worse；　as　in　some　parts　of　Germany　and　in　Russia：　the　State　brought　its　extended　influence　to　bear　in　favour　of　dependence；　and　rendered　commercial　transactions　possible　by　its　protection。　In　most　cases；　however；　the　influence　of　moral；　economical；　and　political　conceptions　made　itself　felt　in　the　direction　of　freedom；　and　we　have　seen　already　that　in　England　legal　doctrine　created　a　powerful　check　on　the　development　of　servitude　by　protecting　the　actual　possession　of　liberty；　and　throwing　the　burden　of　proof　in　questions　of　status　on　the　side　contending　against　such　liberty。　　　　　But　not　all　the　consequences　of　personal　servitude　could　be　removed　in　the　same　way　by　the　conditions　of　actual　life。　Of　all　manorial　exactions　the　most　odious　was　incontestably　the　merchetum；　a　fine　paid　by　the　villain　for　marrying　his　daughter。（48*）　It　was　considered　as　a　note　of　servile　descent；　and　the　man　free　by　blood　was　supposed　to　be　always　exempted　from　it；　however　debased　his　position　in　every　other　respect。　Our　authorities　often　allude　to　this　payment　by　the　energetic　expression　'buying　one's　blood'　（servus　de　sanguine　suo　emendo）。　It　seems　at　first　sight　that　one　may　safely　take　hold　of　this　distinction　in　order　to　trace　the　difference　between　the　two　component　parts　of　the　villain　class。　In　the　status　of　the　socman；　developed　from　the　law　of　Saxon　free…men；　there　was　usually　nothing　of　the　kind。　The　maritagium　of　military　tenure　of　course　has　nothing　in　common　with　it；　being　paid　only　by　the　heiress　of　a　fee；　and　resulting　from　the　control　of　the　military　lord　over　the　land　of　his　retainer。　The　merchetum　must　be　paid　for　every　one　of　the　daughters；　and　even　the　granddaughters　of　a　villain；　it　bad　nothing　to　do　with　succession；　and　sprang　from　personal　subjection。　　　　　When　the　bride　married　out　of　the　power　of　the　lord　a　new　element　was　brought　to　bear　on　the　case：　the　lord　was　entitled　to　a　special　compensation　for　the　loss　of　a　subject　and　of　her　progeny。（49*）　When　the　case　is　mentioned　in　manorial　documents；　the　fine　gets　heightened　accordingly；　and　sometimes　it　is　even　expressly　stated　that　an　arbitrary　payment　will　be　exacted。　The　fine　for　incontinence　naturally　connects　itself　with　the　merchet；　and　a　Glastonbury　manorial　instruction　enjoins　the　Courts　to　present　such　cases　to　the　bailiffs；　the　lord　loses　his　merchet　from　women　who　go　wrong　and　do　not　get　married。（50*）　　　　　Such　is　the　merchet　of　our　extents　and　Court　rolls。　As　I　said；　it　has　great　importance　from　the　point　of　view　of　social　history。　Still　it　would　be　wrong　to　consider　it　as　an　unfailing　test　of　status。　Although　it　is　often　treated　expressly　as　a　note　of　serfdom　（51*）；　some　facts　point　to　the　conclusion　that　its　history　is　a　complex　one。　In　the　first　place　this　merchet　fine　occurs　in　the　extents　sporadically　as　it　were。　The　Hundred　Rolls；　for　instance；　mention　it　almost　always　in　Buckinghamshire；　and　in　some　hundreds　of　Cambridgeshire。　In　other　hundreds　of　this　last　county　it　is　not　mentioned。　However　much　we　lay　to　the　account　of　casual　omissions　of　the　compilers；　they　are　not　sufficient　to　explain　the　general　contrast。　It　would　be　preposterous　to　infer　that　in　the　localities　first　mentioned　the　peasants　were　one　and　all　descended　from　slaves；　and　that　in　those　other　localities　they　were　one　and　all　personally　free。　And　so　we　are　driven　to　the　inference；　that　different　customs　prevailed　in　this　respect　in　places　immediately　adjoining　each　other；　and　that　not　all　the　feudal　serfs　descended　from　Saxon　slaves　paid　merchet。　　　　　If；　on　the　one　hand；　not　all　the　serfs　paid　merchet；　on　the　other　there　is　sufficient　evidence　to　show　that　it　was　paid　in　some　cases　by　free　people。　A　payment　of　this　kind　was　exacted　sometimes　from　free　men　in　villainage；　and　even　from　socage　tenants；　I　shall　have　to　speak　of　this　when　treating　of　the　free　peasantry；　I　advert　to　the　fact　now　in　order　to　show　that　the　most　characteristic　test　of　personal　servitude　does　not　cover　the　whole　ground　occupied　by　the　class；　and　at　the　same　time　spreads　outside　of　its　boundary。　　　　　This　observation　leads　us　to　several　others　which　are　not　devoid　of　importance。　As　soon　as　the　notion　arose　that　personal　servitude　was　implied　by　the　payment　of　merchet；　　as　soon　as　such　a　notion　got　sanctioned　by　legal　theory；　the　fine　was　extended　in　practice　to　cases　where　it　did　not　apply　originally。　We　have　direct　testimony　to　the　effect　that　feudal　lords　introduced　it　on　their　lands　in　places　where　it　had　never　been　paid　（52*）；　and　one　cannot　help　thinking　that　such　administrative　acts　as　the　survey　of　1279…1280；　the　survey　represented　by　the　Hundred　Rolls；　materially　helped　such　encroachments。　The　juries　made　their　presentments　in　respect　of　large　masses　of　peasantry；　under　the　preponderating　influence　of　the　gentry　and　without　much　chance　for　the　verification　of　particular　instances。　The　description　was　not　false　as　a　whole；　but　it　was　apt　to　throw　different　things　into　the　same　mould；　and　to　do　it　in　the　interest　of　landed　proprietors。　Again；　the　variety　of　conditions　in　which　we　come　across　the　merchet；　leads　us　to　suppose　that　this　term　was　extended　through　the　medium　of　legal　theory　to　payments　which　differed　from　each　other　in　their　very　essence：　the　commutation　of　the　'jus　primae　noctis；'　the　compensation　paid　to　the　lord　for　the　loss　of　his　bondwoman　leaving　the　manor；　and　the　fine　for　marriage　to　be　levied　by　the　township　or　the　hundred；　were　all　thrown　together。　Last；　but　not　least；　the　vague　application　of　this　most　definite　of　social　tests　corroborates　what　has　been　already　inferred　from　terminology；　namely；　that　the　chief　stress　was　laid　in　all　these　relations；　not　on　legal；　but　on　economic　distinctions。　The　stratification　of　the　class　and　the　determination　of　the　lord's　rights　both　show　traits　of　legal　status；　but　these　traits　lose　in　importance　in　comparison　with　other　features　that　have　no　legal　meaning；　or　else　they　spread　over　groups　and　relations　which　come　from　different　quarters　and　get　bound　up　together　only　through　economic　conditions。　　　　　The　same　observations　hold