Nowshahr, Northern Lran

Anthropologieet Préhisloiru, 100, 1989,103-1ll

Surnames and gene flow in Shahrestan Nowshahr, Northern lran

Haideh MEHRAI and Eric SUNDERLAND

Abstract

Demographic data were collected to estimate the degree of biological kinship within and between the subdivisions of Shahrestan Nowshahr, using Lasker's (1977) isonymy model. The results indicate some degreeof surname localization within the dehestans and the contiguous regions. The applicability of the model to Iranian data was also assessed.

Résumé

Des données démographiques ont été récoltées dans le Shahrestan Nowshahr pour déceler le degré de prenté biologique, intra- et inter-subdiuisions, en utilisant le modèIe d'isonymie de Lasker (1977). Les résultats indiquent un certain degré de localisation des noms à I'intérieur des dehestanset dans les régions contiguës. Les auteurs s'interrogent sur l'application du modèle aus données iraniennes.

INrNooUcTIoN lated for long periods of time due to its inaccessibil- ity and lack of good communications. Many of the Surnames became mandatory in Iran in 1925, spatially separated localities were inhabited by small when the Civil Registration Office was established. extended families, with marriages confined to the in- The CRO conducted a nationwide system of registra- dividual villages. However, many of these families in- tion whereby each inhabitant was issued an Identity termarried with contiguousfamilies and consequently Card. In conjunctionwith this, the adoption of a sur- formed a larger network of kinship. Behnam (1968) name was also required. definesone type of rural family structure a^sthe "ex- Before 1925 family names were limited to a mi- tended patricentral family", where related families nority. Sometraced their origin in the Prophet's line, are supervised by the extended parental family, and while others were a.scribedby kings in accordance inter-familial marriages are maintained. with their services. The Taifeh (a kinship group) re- Ancestral names have persisted in these areas, ferred to numerous tribes that originally inhabited but their frequency varies according to the size of the the rural regions of the country. They comprised a village, geographic location and the economic struc- network of relatives which originated a.s an a^ssem- ture. In some instancesdouble-surnames appear, in- blage of families linked by geographic proximity, or dicating the union of two families. In the present loyalty to a common chief (Behnam, lg70). Due to study, the samples from each village are too small isolation these families eventually intermarried, and for analysis, therefore names have been analysed at the kinship group was enlarged. 1'he Hezerfamil re- the dehestanlevel, which necessarilyreduces the sig- ferred to various hereditary families who numbered nificance of the extent of localization of names. Ad- one thousand. When surnames were adopted, the ditionally, it would be very difficult to observe dis- choice of name was based on various factors. Some tinct clines in surname flow between the dehestans, chosethe name of their village, while others followed becauseof the irregular distribution of the villages their trade, e.g. Khatat (scribe). The suffix "zadeh" within each dehestan. In general,surnames are not (born to) was added to some forenames,€.g. Hossein- very mobile due to the low mobility of the population "y" zadeh;others added the letter to the name, im- itself. plying belonging to, e.g. Rassouly. Sincethen, names One means of examining the degree of biologi- have been regularly transmitted from the male parent cal kinship within a population, or between its sub- to the child. Since birth registration is mandatory, a divisions, is to measure the occurrencesof common child takes the name of the father. Illegitimate chil- surnamesusing Lasker's(1977) isonymy model. The dren are usually not given a name unlessrequired by method of measuringgenetic lineagethrough isonymy either parent. The wife usually takes the name of her is useful in studies where information on kinship husband at marriage. based on pedigreesand genetic markers is not avail- A large part of the Caspian littoral remained iso. able. Surname transmissionresembles genetic inher- 104 Haideh Moxut and Eric SuNoeRl.nNn itance, and ca,n be analysed to indicate the origin of random (attributed to population size and surname genes, and to delineate breeding populations. Sur- distribution), and non-random (special proclivity to names ca.n become extinct, like genes, depending on marry one's own cousin) components of inbreeding. name frequency, family size, and sex ratio (Dobson Lasker (1977) extended the method of isonymy from and Roberts, 1971). intracommunity inbreeding to inter-population rela- Communities with few surnames at high frequen- tionship, to show how it was possible to examine the cies indicate greater localization of genes, a,nd con- relative linkage of two or more populations by sur- "Coeffi- sequently high inbreeding coefficients. Estimates of name analysis. He introduced the concept of rates of inbreeding can be obtained through mar- cient of Relationship" by isonymy, as the proportion riages between persons of the same name. Crow and of individuals with common surnames within or be- Mange (1965) elaborated this idea, and examined the tween populations.

Fig. 1 : Map of Iran and the administrative subdivisions of Shahrestan Nowshahr. :v;"$!!,,.,, i-:l-t',;*èù'-' ..-cÏ ! J Nowshalir / -:J'-ù..- t.. n--f'---.r"': AFGHANI S T AN - --1--;'- | R-lÀrrN^ ,1- r )-t r-(. ^('-

PAKISTAN

SAUDI ARABIA n :A\à"

5 10 15Km Caspian

.tl 'o. Sea :g,nurï; ! a a, a ata KACHROSTAGH (t,, a I iil:, ÊI{e$aramte 'Hagsank,"J, ' .t '.,1' d'U/-Kts'N^>t ;?Éar-1eEH-.i >i t t ..'F vJf' , K|:l-tôlrR / u\ t j ? -_j_4î--:7i1. $:fi#*.ti:"4'!*o;;.h,. ru,..A.. .(.-aarj\l* ,'i ^.:;\ . .. ,T.-..*\rrr;ir_ 1. o o tiùr#- __., \) e. ra .*...,1..' io"$^arl H - \ .touo:.:-.': îS KU L' . ,KUHEsrm.t xo.rôun. I $ r\ 1J"-. ,i*r."tsuù 1 \ 7 r' I qr. S e-sxnh rq.. f"iBAKHSH Bakhshboundary . r loenesunDEHESTAN Dehestan boundary - - - Towns Towns 0 Villages. Surnames and gene flow in ShahrestanNowshahr 105

S:rUoY An,EA localizationexists, and to examine any clines in surname flow between every pair of dehestans. It was also Shahrestan Nowshahr (fiS. 1) is an adminis- considereduseful to assessthe possibility of applying this technique trative subdivision (3695 sq km) of the Province of to an Iranian population. Mazandaran,situated in the North of lran, between The coefficient of relationship by isonymy is the southern shoresof the Caspian Sea and the Al- mathematically expressedas borz Ranges. Administratively, Iran is divided into ostans, shahrestans,bakhshes, dehestans (conglom- Ri - S;z)/2NtNz eratesof villages)and dehs (villages). The deh is the f{Sn smallest unit of habitation, urswell as the economic where ^9;1and unit of agriculture and consumption; it also consti- .9;2 are the respective frequenciesof the ith surname in the tutes a social unit composedmainly of related indi- males and femalesof the population viduals. The great majority of the rural population (or in two populations), and N1 and N2 are the corresponding of the Caspian Littoral residesin these units. These sample sizes. The model is based on two assumptions villages are semi-continuous on the coast, and nucle- : unilineal descentof surname (polyphyletisnr or multiple ated in the highlands. The climate is subtropical in ancestral origin of sur- namescan violate the lowlands, and is one major causeof the summer the assumption),and that the ex- changeof surnames proportional and winter migration of the population. The moun- is to the flow of migrants, i.e. equal tainous topography and the abundanceof rivers have virilocal and uxorilocal migration. constituted major obstaclesto movemeut and contact The population under study has a system of pa- betweenthe villages. The construction of roads and trilineally descendingsurnames; furthermore, it has bridges, specifically the Chalous bridge have facili- been assumedthat surnamesare monophyletic. The tated communication between the settlements. The assumption of uniform migration of the sexes has inhabitants are mainly Mazandarani and Gilaki, in been consideredwith caution. In general,patrilocal- addition to the Khajevand and Lak Kurds who were ity appears to be more frequent, and more of the ge- resettled here around 200 years ago from north-west netic exchangeand intermarriagebetween the settle- Iran. Many of the local populations are culturally ments are basedon female migration. However, some homogeneousand retain strong kinship ties. factorsneed to be considered.Primarily, movementis The comparative demographic analysis of the extremely limited becauseof the geographicsetting, three existing censusreturns of ShahrestanNowshahr the communicationnetwork, the agricultural pattern (Mehrai, 1984) has revealedchanges in the composi- of subsistence,and also the proclivity to marry mem- tion of the population, and the trend conforms to bers of the same village. It is erroneousto generalize on pattern the demographicgeneralization of the tansition Pe- the of movement within each dehestan, riod. Differencesin family structure, migration pat- since it is determined by the structure of the distri- terns and population growth betweenthe urban and bution of the villages,the kinship structure,and its rural areas, have rural or urban nature. Another factor is the local cus- been attributed to socio-economic "khaneh-damad" changes,and urbanization in the last 40 years. tom of meaning house-groom;this is a situation when a man movesto a wife's residence at marriage. Touba(1972) statesthat matrilocal res- M.mpnTAL AND METHoDS idence also occurs dependingon the village and the economicstate of the bride's family. A final essential Data presented for this study were collected point is the traditional summer and winter migration through questionnaireforrns, from a total of 1,042 of the people,particularly the males who descendto households. The forms were distributed to schools the lowlandsin the winter in searchof work. and usu- throughout the region, to obtain a geographically ally migrate back to the mountains in the summer. random sample of the population. Schools in the A proportion of these local migrants remain in the Shahrestan are limited to large villages and urban lowlandsbecause of better living standards. centres,and consequentlyevery village could not be These reasonsprovide the base data for the anal- directly sampled. The data extracted for this study ysis of isonymy. Estimation of the rti coefficients were based upon the names and birthplaces of the are basedupon 1293patronyms and matronyms (652 respondents'parents. malesand 641 females),who have been born and re- The purpose of the present analysisis to study side in the nineteen subdivisionsof SharestanNow- the data on the frequency and distribution of sur- shahr. names,and by applying Lasker'smethod, determine Cluster analysisand non-metric multidimensio- how geneticallyisolated or panmictic the inhabitants nal scaling (Kruskal, 1964) are employed to demon- of each dehestan is, whether a pattern of surname strate the hierarchicalpattern of relatednessbetween 106 Haideh Mngn.rt and Eric SuttopRr,lNn the dehestans in relation to commonality of surnames 6.9 Toof the total surnames. Other related surnames and the geographical properties of the environment. are Faghieh-Maleky, Faghieh-Marzban and Faghieh- The analysis was performed by coding the data, and Nasseri, signifying the joining of the Faghieh family transforming them to the computer (Numac, Com- with other families. This dehestanis a large territory, puter Unit, University of Durham). and is inhabited mainly by Kurdish tribes.

RpsuI,ts AND DIscUssIoN The Delfan family, originally a subdivision of the Lak tribe, also reside in Kelardasht, in addition 1. Surname to related families of Delfan-Azari, Delfan-Hosseini, Delfani and Delfanian. The name Delfan also ap- An analysis was carried out by compiling a list pears in Kuhestan-Gharb and Kelarostagh. Sur- of the incidence and frequenciesof surname in each names derived directly from the Lak tribes are Lake- of the subdivisions (some have very few samples) of layeh, Lakpour, and Lakourej-Manssouri, which are Sharestan Nowshahr. It was possible to examine the distributed in Kelardasht, Kelarostagh, and Biroun- preponderanceof specific surnames particula^rto the bashm. The surname Manssouri also appears quite dehestan. Table I presents a rank order of occur- frequently in Kela,rdasht, and Birounbashm. Other rencesof surnames in each unit for both sexes. Gen- frequently occurring surnames are Radaii and Vissy erally, unique nannes(names that occur only once) which comprise ll.9 To and 6.6 To of the total sur- are the most frequent in all the units. The Dehes- names respectively. The family names of Sam, Sam- tans of Kalej, Kelardasht, Baladeh-Kojour, Zanouss- Daliri and Dalir are distributed in Kelarostagh, Bi- Rostagh and Birounbashm, show occurrences of par- rounbashm, Kouhestan-gharb and Chalous City. ticular surnames at high frequencies. flom this analysis it becomes evident that cer- a) Bakhsh Markaai tain namesreflecting family and tribal lines are local- In Dehestan Kalej, the most frequently occurring ized in particular dehestans and distributed around surnames are Divasalar (22.6 %), Salarian (20.a %) contiguousareas, thereby showing some degreeof sur- and Salar (10.7Yo).lt is hypothesizedthat there is an name flow. inter-relationship between these names with possible similar ancestral origin. These surnames comprise 53.7 To of the inhabitants, demonstrating the extent of intermarriage and inbreeding in the Taifeh of Salar. 2. Surnames as genetic markers The Taifeh of Kia, in addition to the related families of Kiakojouri, Kia-Lashaki, Kia-Mohammadi, Kia- Heyrati, Kia-Mansouri and Lashaki are the major The averagecoefficient of relationship (nr) within inhabitants of Humeh, particularly in the village of each of the dehestansbased on the surnames of the Kordkroudsar; they are also distributed in Nowshahr males and the females, and between the dehestans City. In Kalrudpey and Kheyrud-kenar, the Darvish based on the pooled results, is presented in table 2. family reside, while in Kouhparat, the Mataji family Some of the coefficients are inflated because of small are settled. sample sizes, the inclusion of members of close rela- The Khazaii family constitutes the main inhab- tives living in the same household, and by duplica- itants of Zanouss-Rostagh, comprising 23.3 Voof the tion of surnames, if female married names are given population. The name also appears in the contiguous instead of the maiden names. These factors do not dehestans of Baladeh-Kojour (23.8 %), Kheyroud- show the actual isonymic relationships. In general, Kenar (13.8 %), and Nowshahr City and Kuhparat. large sample sizes are needed to yield values of Ri The Khazaii family are one of the numerous families which are not dorninated by stocha^sticfactors which belonging to the Khajevand and Lak tribes who orig- invariably affect small populations. inate from the Zagros regions of Azarbaijan, Kur- distan and Luristan, but were resettled in Mazan- The average .8d within each unit ranges from daran. Within Sharestan Nowshahr, their domains .00735in Kelardasht to .06330in Kalej. The mean lo- have been in the mountainous valleys of Kojour and cal Ri is around .03290,but excluding Alavikola, Cha- Kelardasht (Rabino, 1913). One of their main strong- landar and Kuhestan-Shargh(which have inflated the holds ha.sbeen in the Yelagh (summer quarters) of coefficients),a lower mean of .02695is obtained. The Poul, presently the capital of Zanouss-Rostagh. coefficientappears to be very much a function of the size of the dehestanswhich vary greatly, and also the b) Bakhsh Kelardasht and Chalous number, size and density of villages in each dehestan, The Dehestan of Kelardasht, is another strong- and the structure of their distribution. Another in- hold of the Khajevand and Lak tribes. It is partly influential factor (which cannot at this level of analysis habited by the Faghieh-Abdollahy family, comprising be examined) is the number of taifehs residing there. Surnames and gene flow in Shahrestan Nowshahr 107

È-(OôrÉôr æ(o(O$O Oo)o) O 6t\ rô @s(oæ (o (r) Vç rô| aN ÉS *(V) d) çN È- (o ôl t- È'- s ôl Fl ôl (o o

I ;tÈ-(D6O SôlrON(o F{ôtN (O ço) 'q oæCo(o F) ôl È- !.!. r-.ôt ô|CO Ê.< Hôar{ ô r-a b- (Y) È.. !. rô ôl Fc{(o I

q?QÈ.Q(o QOlO(?)s stôÈ- cr) @(r) rôN È- rô<)tôæ |o c-|6l c{NF{Nr(t ÉlâF. 6 (oç o æF{çt'- ôl I ôl F{ (o E

a o b'-NO)Ê O) t$dlOlô () z EE'RS $B=RER6F E $s s (o E lô rô CD r{ lô c{ o ôl

ô|@ È- c{

Io N ç ôl

ôl .A q) o I N q) o) Èr ()I @ o

@ à C)

lô C' c) a Oco ë o ! e't o) & u u (d ôl L1 th q) o (Ë

o) a î{ È

m td. ts ôl r l.ô

F. CO o l.-

c{ r-{ r-l t-{ F{ .ôl b.- æ :

rô a) S F{FlF{ t- r-a (o C{ cf)ç

(Y) o, ôrôr o) r-t T ô|S6I ôl t'- c) c{ rô OIô o) ro CO ôt: :1. b.-

(O lo (o rr t-- o)o) ôl ôt ooo o coo b- Oo) rô (Y) (o câ ôl (Y) rô N F. ç rôç @ Crî co rô CY) 6) N ôl t.-

oo oo € ugBa tigt o a j Itbt .lo Ê :Ë À i.Ë Êan dlF HËràÈË e 6 F .a, lt + +gFg Ë+È Ë _EgEÈ ËëËË .âO .!i S Ë .rY i (t) .t!é .-- tJ É i47,i 1,â à .r,Zc', ËËsrËË iggg:rïEs t$3 I f legËI Fr 108 HaidehMnHRlt and Eric SuNnpRlaNo

a o q N \ \ na s

o"(ov $ \r'O I v ë8

4"" tôc\\ (â \)oH I X 888 BSRB sA ÈESS () ù sËBËs s3888ç Ë X ËsËËBs

sÈ qËËËËËË FSR8383E b * BSËËËSs$

È F ËËEEËÈBËqt; s 'çào è ËRBiÈsÈBËÈ(â 'ÈJ' s 3888 'èA \ R*-sK Ê Ël (â g SËÈEdBËËBËËq) "t R-R8-E +) 'ès Ë Ëë!ÈËdEÈBEËË(D P G àrs s :3 X ËSËEEÈÈÈËËSËBo s e s EËBRFRÊRFRFRRRÈE qËËÈËsËEssBËEËo s $ Ë Ll \J q) Ë s8ÈÈËËstË8ËËËËÈ?h U .a >{ EsFBisBBBid 'êÈ qqHEëEHsËësÊËsËË.C g at) 'Èo IJ = B \l qËEÊEBËBËSBSsÈEËËF] s P ç- o Fi g qqËËËEËËËËËËËËqBËq{s s a) 'Fù ESSSFASFSBSEEEFç8çR$ b -ra s ËlËëEËësësËsëËSsËsBq) I "i $ tsJ "a iFËuHÈ$$$RRiÈ$T$iiË* Surnames and gene flow in Shahrestan Nowshahr 109

o.ooo5

o.oo30

o.oo60

o,oo95

o.o125

o.o155

o.o190

o,o220

o.o255

o.o285

Kal Bal Zan Kuh Cha Khe Tav pan N-C Hum Kpy C-C Ala Kld K-E Bir Kel K-W

Fig. 2 : Clustering of subdivisions based on isonymy.

Ktd Tav

Khe Cha Blr Bal

+ Zan c-c Kuh K-W N-C K-E

Kel Kpy Hum

Fig. 3 : Spatial positioning of the subdivisions based on isonymy. 110 HaidehMBHRII and Eric SuxopRraNn

The dehestans of Kalej, Baladeh-Kojour, Za- of isolation, and a low coefficientof relationship with nouss-Rostaghand Birounbashm show relatively large other communities. It appears that it would be more coefficients. The high values represent the homoge- instructive to ignore administrative boundaries, and neous composition in terms of common surnames, to consider actual distancesfor ,Rdestimates. and the accumulation of relationship through relatively high levels of kin-endogamy. The two cities of Coxcr,usroN Nowshahr and Chalous exhibit lower values, reflect- ing the heterogeneouscomposition of the inhabitants The analysis of the frequency of common sur- and surnames. The dehestanof Kelardasht showsthe names and the extent of biological kinship within lowest coefficient, and this is attributed to its large and between the dehestansof Shahrestan Nowshahr, size and the existenceof different tribal groups there. has provided information on the pattern of migration and the direction gene The ^Ri coefficients within each subdivision are of flow, and has indicatàd that some degree larger than those between them, indicating some de- of surname localization exists. Although informative patterns gree of localization of surnames and genes. The emerge from the analysis, i.e. a geographic pattern mean coefficient of relationship between the dehes- is seen as in the correlation (although tans yields a value of .00150,ranging from 0 to .0304b. low) of .Ri with distance, and in the hierarchi- cal structure, The standard deviation is .007, indicating the high there are limitations to the inferences variability of the coefficients. to be drawn from the results. The observed coefficients should be consideredas relative and not abso- The largest between ,Ri coefficientsoccur in the lute indices of isonymic relationships. Becauseof the contiguous dehestans of Bakhsh Markazi, in which short time since the origin of surnames in this region, they form small and large clusters in terms of com- the random element of inbreeding relating to the ac- mon surnames. The dehestans of Baladeh-Kojour, cumulatedhistoric aspectcannot be directly inferred, Zanouss-Rostagh, Kuhparat, Kheyroud-Kenar and and therefore conclusionshave been very limited. Chalandar form one cluster. The .Ei values between There is a positive correlation them range from .01095 to .0304b. Two small sepa- between the pattern of migration based on parental rate clusters also appear : the dehestansof Tavabe- mobility (which is a better index of population structure), Kojour and Panjakrostagh and, secondly, those of and that ba^sedon surname analysis. Both Kachrostagh and Kalej; both link to the main clus- models have sub- stantiated the observed structure pattern ter. Other dehestansshow far smaller ^Edvalues, but and of migration, demonstrating the primary the remaining dehestansin Bakhsh Markazi are the subdivision of the Shahrestan(historically first to unite with the main cluster, followed by those maintained by the Chalous River), and the continuing closergenetic in the Bakhsh of Chalous and Kelardasht (fig. 2). relationship maintained within each division. Figure 3 distinctly shows the positioning of the dehestans by commonality of surnames; it very closely Shahrestan Nowshahr is characterized by gen- resemblesthe geographic positioning. More impor- erally low mobility of the population because of its tantly, it demonstrates the manner in which the two land-based economy, and a tendency to virilocality, main clustersemerge separated by the Chalous River, and a propensity to patrilateral marriages. These which has historically been a dividing line within the factors will tend to augment the isonymy indices. In Sharestan. employing the method of Lasker, we intended to examine its applicability to Iranian data, but the main The coefficients between the subdivisions vary emphasis has been on extrapolating on the mating according to the nature of the regions. There is some structure and the biological kinship within the sub- geographic element inherent in the distribution and divisions of the area. The results from the surname flow of the surnames. In accordancewith Malecot's data and the migration data correspond very well, (1967) theory that kinship between populations de- and a consistentpattern emerges. We hope that fur- clinesas distanceincreases, a larger .Bris expectedbe- ther researchwhich is in processwill clarify some of tween contiguous arein. A small but negative corre- the deficienciesinherent in the results. lation appeared between ^Edand geographic distance (. - -.211) indicating that the heterogenousdistribu- tion of the settlements, local topography, and the kinship network are also influential in restricting movement. It is evident that the non-random flow of sur- References names causedby the topography and the cultural re- BoHruau, J., 1968. Notes towards a typology of tran- straints would result in partially homogenousdemes. sitional family forms in Iran. In : Amani M. Lasker (1977) claims that a low correlation is an indi- (ed.) : Some demographicaspects of the popula- cation of the Island Model of genedistribution, where tion of lran. Institute for social Studies and Re- each population division demonstrates a high degree search,University of Tehran, Tehran. Surnames and gene flow in Shahrestan Nowshahr 111

BEUN^lu, J., 1970. Nuclear families and kinship ogy,49: 489-493. groups in Iran. Diogenes,T6 : 115-131. Ma,lBcor, G., 1967. Identical loci and relationship. CRow, J.F. and M.tNce, A.P., 1965.Measurements In: Lecam L. and J. Neyman, eds. : Proceedings of inbreeding from the frequency of marriages of lhe fifth Berkeley Symposiurnon mathematical between persons of the same surname. Eagenics slalislics and probabilily,4 : 317-332. Quarterly, t2 : 199-203. MottR.lt, H., 1984. A demographicstudy of somepop- DonsoN, T. and RonoRrs, D.F., 1971. Historical ulations in Shahreslan Newshahr, Mazandaran, population movement and gene flow in Northum- Nodh Imn, with rcfercnce lo the genetic stttrc- berland parishes. Journal of biosocial Science, lure. Unpublished Doctoral Thesis, Dept. of An- 3: 193-208. thropology, University of Durham. KRusx^e,1,J.8., 1964. Multidimensional scaling by R.nntNo,H.L., 1913. A journey in Mazandaran(from optimizing goodness of fit to a non-metric hy- Rasht to Sari). Joarnal of the Royal geogmphic pothesis. Psychometrika,29 : l-27. Society,42 : 435-454. LrsxeR, G.W., 1977. A coefficientof relationshipby Tounn,J.R., 1972.Marriage and thefarnily in lran. isonymy : a method for estimating the genetic Institute for social Studies and Research, Uni- relationship between populations. Human Biol- versity of Tehran, Tehran.

Authors'addresses: Haideh MpHul Kordkroudsar 40 Modares Road NOWSHAHR 46516(Iran) and Eric Sunderland University College of \{ales BANGOR, Gwynedd LLï 72DG (G.8.).