In t_his pape1 we analyze the structure of thelibrary murketfor.s:ientific and technical (ST) serials. The anahlsis takes the fonn of un exbrci,se aimed at u theoretical reconstnrction ofthe ST serials ho[dl."gt n7 fSU Libruries after almost a rlecutle of massioe cancellations antl u policrl of a&Iing no niu subscription's. This exercise toas done in conjunciion rtith the Louisiana state Projeit (SRP), ancl it utilized an experie Serials Eouluutor. Much of the poTter is initions, meusures, and algorithms nece.sil,:'# "' ;ff: #::,: jJ :r?#' ;.," e, an,t u; e , he strengths andweukni,sses of faculty rutings Chemi.stry pluyedthe role of the test discipline, and other ii flelds"uere inacstigured to tlelermine whether the protess'es uffecr.ing chemisiru are ul.so uctioe inthem. We deoelop thehypothe,iis that humun knowletlgefunctions on the sume_prypabilittl stnrcture is biological nature und societll.' iVe show that this probubility structure results in the highlq skeued, stabie distributions thul churocterize the social strdificution ,sqste"m of science ttnd technology trs wcll rs of the serials sqstem bu.sid upon it'.' science andtechnology are seen inthis paper us dominatedbq stable elites, who tentl to center rround IratlitionuLlq pristigious institutions ttnd Ttublish their work in IJ.s. ussociution journals.'Consequentlq,IJ.s. ttssocitttion sertrtls huue higher sr.uulue, urultheq plaq tt tlomin'ant roie not onltl in internallihrurq use hur also in i.nterlibnr;i lod;. Due to their higher ST o'tilue, U. S. zrssr> ciution journals can be sokl to librarie.s in greuter numbers at cheaper pri,ces than the jountuls of commercitLl publishers, anrl thi.s carLses the'sr ierluts mtrkel to bifurufte, with sr aulue tantlingto concentrtfte onrhe u.s. ussoci.-tion serials ancl costs on the commerciul ones.As a resub of thehlghly skewed, stable nuture of the ST serials sqstem,the