.1 iu jy CLI,- traini--ri matl,oes arn based on tzclln4-ttes (as -ty cor.Ll;f,.rnd ty-'tli ecc;lcoical and oti-,c-- Llisci.-line3. devictek,7@ frct:-, -.hc.. clissi,.al arul a r 4-an a@i-,d p--vloviar., i?ip-ws bccau:;tt of 6-a-- reco@.7-i-40- 04-' tht: cor.-- n:ad irt.-ort-"nce of the bet!,?een 'Anterrtal a---d external 2 cocr.,,.'Lze even ttie si-wpler;* cond4tfor@S rct, .11' -.. L- - k := y a i-@ore 0'!- less co.T,.nlicated c-tiaivi or sequ,..-:zice. to cituaticn* i-- cr-c of dec; nte fc.!l stil.,ir-c-s 2 ten@'. to bt-- of tlztf?. go-n,-) t, Req, oul@ject as its L-,7,-i 4-:Or data arc LeFt to a co=-,q,-.surate v?'6t:li 2 v--, fca"- to tli,- succes--@ of tl,?e Cur c-n--@-aieer-;.n,T J-'s I.ascie-, o,-- @ro- oLr of t.ie in -ad en situation, CeZree 2o' ra- quired, equl-pr,,eut avz:41.;-a6le and c.,>sL- cf t-i-,,e fa--torr,. The trp-irLiz--- plwa is ta4-loree- to @14.t tl-..e -reru-'Lrome-,itr,-2 of thL- -s or rroba:r,'e i-,rollen- areas of Cr-4.tic" the proposra 1--al-ai@iorr.1 ar;-- and a.-i-r@iie,- o-@r rp- 2 proacl,, are :L-aitiatee- rrct-ir:T.-- of a cr--ic-I nature we tr,- to a-@--tEcit or, 7nore cre fro-rt T'le c!a-@a rec-iirements are -Imclu,-:ed. i-"- t!@.c -@--a-A@.ni C, Our -rainin@- s rt es-ones bu-, the milecto-ties L. :, pla:-. a"-so !.ziclude .4..L seldo.-- hive dates at-lac!iE-4. A't -Is to ircdicl2@- ou-i-ie L--Leer 1--i-C o&' rzte'l o--F poor coztrol oAz occur oi@--siec the labcrato-nt. 7--.a2-in-Lng 'Ls a -rovti-. procees ar-@4 to ptic',i ali@:,:a Cr, Z f'LY.Cd CCiIC-L'-JIL i.,--;titrut r;ai7.::a; a is a.-, un-.:i--r-- -,rzzctics an,-@ cour=,-, .Call-la Two Daily trainina pro-rams aro'const,-uctcd for each anira-al based on his prev.-Lous record-particularly, #.he pre-,rious day's re-ord, bu&. 121so usin- t'&-Ae carlier're-ccrds of rate of proi,2,rers, litc'.dence of re-ression, health conditions, ttc; Occasionally rcu%-ine steps of progress svich as incrczc4-ng range a certain nul-ber of yards are sacrificed for increased accuraci c-A" d4-s--rir,-ination2, pol-.'Lsh- i-,ig certain skilled torrinal bch,@viors, corrrc-@'Lnc-, or ci.-t-A.U,-uirh- in- problem behaviors and the Verj- little quantit.-tri-%,,e or qiiil-@Itptive e-.,qzperimental evidence 2ezdsts on lon- cliaini: of behavior such as arc ir.-rolved in the present project. Tnere are 1--'ttle data avaij-lable on extenzibil- it-j of EecoDdarj reinforcerent over a'.Ion" time periods of hif,,,n e4L@'J'ort ta2sk.-. The ev4&dence --va4-lt7jle noiz, !,.or7cvdr sli@ht it ray be, ru-zt fo= the basis of our present design @nd we must gather the necessary data for future entinearir.- designs. In vie,,i of the ra-ic@4-ty of pe2r&-.ine-.it evidence a carj-ain at,.our-.t.;of art, sl,7111, eve'Luation and edu--ated cues3-vor'l.-. ru----,-- be used in judg-,'.r,Z !-.ow far can be rtretclled eaci-, da:.-, i?hen ne2A,7 s4-t-aations cw.-, be introduced anL' behavioral can be dropped without disruntin- tre L From Septe-a@,er *..%rou@h Decezber o-F 1967, we P-nal7zed the prog-zam goals a-ad requ'-rericits. 2 Ttic roal-s r.-cre to get t-he val-,-Lcl-a to c remot,-- poirt and b ar-1; a- A s--mgle te=inal be-Iizv4-or r;.-ar, be required. ',lie -uidance r,y!:ten urist util-izc tile capabi -Ltie2s of a-,i r... F In our arnlys-As, t-e7o areas seered to be viost cr4-tical ard in rleed o 4' -Ln,.,es-igation, develop-,a-:!nt as-id selr--c-z:-Iott o-r guida--Lce -Ctl-l"-l 2 3 a-,id es-.ablichin- a mr-xi=uT-- -ange cap-37r;-,lity. We proposed several a-proaches to t.ie guidlmce problem, one of r, which was not directly comzat2i@ile ttie R r Li-,il@. '.'.IC proposed to assess che cf-.'&'Icacl OZ Continuotis :Ln,-Ior.:at-'Xor. in-)ut svz;ters a:id inn-.,,ts syste--s an u!cll as com7pare 9 poz'-tive infor---z:ion c,,Gtem- 5yzte-,ns. ra'ge T'L re e- Once the system(s) of cicice had i)ecti selected, i-ic- p-roposc-.d to gather data on raa-e e-.-.tension. I'@e wcre interested in'..biologi- cal fzntors sucli as -.ttc-.2tion rpan,2 pliysical endura.--ce, total raiige aiid ej'fcct of different reinforce-Tc-nt sciedules. Onca i.-.e had a sufficient a-.-cunt of data to stipport the sys-- tan could po5sibly tchieve the desired ran-e we propo@-ed to 2 CZ incor7,7o-@ate 'r)cta-.,ior-@ 2rd -,, t te@-,,) %-. to a- ciiie-,7e rax--'min rerrc. 'Ln--ludc--' at tl-,is - time a'Lso would be the @-ystcr,s lrt,,t;@ration asrects. A) 'ilie training protocol involves syste,--aticnlly in-.reising the co,-i-)Icxity mid slc4@ll level of the aniral responses under the conditions ot4 an and '--cquently 6--an-in-2 env4-rorment. l,'e conditio-ned tlie ar-izrals to se.-.reii for tar-at-- i2.-iL' re@-pond to L; the targets once tl--,ey tliein. 1;e supplied aud@, tory cues to the an2imals to aid thei-- in; their searc'ii for targets. 1-!-e also attem,,)ted to train an a-.iimal iiit;.icut the use of %-,irgets but the res-alts in comparison to 'the tarfat siturt4t.O'n T,7C!,re poor. T'i4a au(:titory cues t;ere a conti-.i-uous sienal %.7'ien the animal was haaditt- towards the i)roper tar-et. a si-nal ireanir,- turn to the left (target on left) a-.id a si,--nal i--ezning turn r-J.&4t (target on Tiiese si-nals -t:ere used in differ2E-.-.it ccv-j-.-Lriatica @gor C .di-fferent ani,-k--ls dependin- or, the parl.--A@.cular annroacl-. in -Lrain-- C2 in&. The bas-A.c appro@-c'Aies -;,2,ere2: 0-,i cr-,urse Lcf- I.-/ 1:,ig IL @t On course--Left/T".i(c@.I.It .@rnether the --uidance signals were continuot@s or interr"-Lttant was another varial-@le in2 our tr-ain4.n; schexe. The 4@nit'Aal tra'Ln.:Lng is accoRDlished xty4-tii the relatively sterile confines o'L a sr,,-ll )pen--essentially an oversized skinner box! Ouce the br-lia,.@ior -,..,as established in the sp-,al@l pen we roved to a l@,r-er pen wl.tb a rore complex envircn=enf-- P.-ad also intro%'Auced -ate tra4-nin,-, atid rorc extens4.'Ve her.,-'lirg procedures. By thir. tietliod of coitiniial euviroriver-tal and I)th.-a-,riorE4.1 e>.-pansicn %,,,e roved the a,.ii=als outside the cor?cunds. Par.,e rru.- 71-Lc- datz-. oA@' tti(@ trp-'A.i,.in- w--re c:- trials, posj4&-ion-- ran oF. tcrr Is;, nur--(.@r o4:- corrcct cl.-.oi2cc-c, .ors fac* n t i= aitd :4-t zal tir.L-s, c.c--ta-Alleci r,--cor"' o'@