Astheaboveshows，weknowpracticallynothingdefiniteaboutthefamilyandothersocialgroupingsoftheanthropoidapes。theevidenceisflatlycontradictory。Whichisnottobewonderedat。Theevidencewithregardtosavagehumantribesiscontradictoryenough，requiringverycriticalexaminationandsifting。andapesocietiesarefarmoredifficulttoobservethanhuman。Forthepresent，therefore，wemustrejectanyconclusiondrawnfromsuchcompletelyunreliablereports。
ThesentencequotedfromEspinas，however，providesabetterstartingpoint。Amongthehigheranimalstheherdandthefamilyarenotcomplementarytooneanother，butantagonistic。Espinasshowsverywellhowthejealousyofthemalesduringthematingseasonloosensthetiesofeverysocialherdortemporarilybreaksitup。
Whenthefamilybondiscloseandexclusive，herdsformonlyinexceptionalcases。Whenontheotherhandfreesexualintercourseorpolygamyprevails，theherdcomesintobeingalmostspontaneously……Beforeaherdcanbeformed，familytiesmustbeloosenedandtheindividualmusthavebecomefreeagain。Thisisthereasonwhyorganizedflocksaresorarelyfoundamongbirds……
Wefindmoreorlessorganizedsocietiesamongmammals，however，preciselybecauseheretheindividualisnotmergedinthefamily……Initsfirstgrowth，therefore，thecommonfeelingoftheherdhasnogreaterenemythanthecommonfeelingofthefamily。Westateitwithouthesitation：onlybyabsorbingfamilieswhichhadundergonearadicalchangecouldasocialformhigherthanthefamilyhavedeveloped。atthesametime，thesefamiliesweretherebyenabledlatertoconstitutethemselvesafreshunderinfinitelymorefavorablecircumstances。[Espinas，op。cit。，quotedbyGiraud-Teulon，Originesdumariageetdelafamille，1884，pp。
518-20]。