HARVARD UNIVERSITY Library of the Museum of Comparative Zoology Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. XCV, No. 1 REVISION OF THE AFRICAN LIZARDS OF THE FAMILY CORDYLIDAE By Arthur Loveridge With Twelve Plates CAMBRIDGE, MASS. U. S. A. PRINTED FOR THE MUSEUM November, 1944 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE The Bulletin and Memoirs are devoted to the publication of investigations by the Staff of the Museum or of reports by spec- ialists upon the Museum collections or explorations. Of the Bulletin, Vols. I to XCIV, and Vol. XCV, No. 1, have ap- peared and of the Memoirs, Vol. I to LVI. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately- A price list of the publications of the Museum will be sent upon ap- plication to the Director of the Museum of Comparative Zoology, Cambridge, Massachusetts. After 1941 no more Memoirs are to be published. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. XCV, No. 1 ^REVISION OF THE AFRICAN LIZARDS OF THE FAMILY CORDYLIDAE By Arthur Loveridge With Twelve Plates CAMBRIDGE, MASS. U. S. A. PRINTED FOR THE MUSEUM November, 1944 No. 1. — Revision of the African Lizards of the Family Cordylidae By Arthur Loveridge CONTENTS Page Introduction 1 Index to the Species recognized 6 Key to the Genera 8 Bibliography 110 INTRODUCTION That the name ZONURIDAE should have been employed by Boulenger (1884a), was most unfortunate, for Zonurus Merrem (1820) was antedated by Cordylus Laurenti (1768), as has been pointed out by Stejneger (1936b). As article 5 of the International Rules of Nomen- clature requires that the family name be based on the type genus, CORDYLIDAE must replace ZONURIDAE, however much such major changes are to be deplored. In 1885 Boulenger recognized the family as comprised of 14 species, as against 48 forms enumerated in the present revision. In 1930 Power, when preparing a key to the South African species of the genus "Zonu- rus", listed 18 members and, though two of these {capensis and robertsi) are here transferred to Pseudocordylus, no fewer than 28 species and races of Cordylus are accepted in the present paper. Apart from desul- tory comments, no revisionary work1 has been done on Pseudocordylus or Platysaurus, but Boulenger (1899c) furnished a synopsis of the 7 species of Chamaesaura, now reduced to 5 forms by synonymy. The present revision, like others in the series2, is an attempt to furnish a synopsis of all additions to our knowledge of the family since 1885, and a serious endeavour to arrange its members according to probable lines of evolution. In this connection I entertain doubts as to 1 The present revision (completed in the summer of 1942) was just going to press (1944) when a copy of FitzSimons' (1943) "The Lizards of South Africa," was received. The two new races of Cordylidae which he describes have been included or discussed in this revision and a few footnotes and comments inserted, but his book must be consulted for other data, English names for most of the species, and many fresh locality records. While we are in sub- stantial agreement as to the entities of most forms, my phylogenetic conclusions are so widely divergent from the arrangement devised by FitzSimons that it must be left to others to decide which more truly represents the probable lines of descent. 2 Revision of the African Lizards of the Family Amphisbaenidae, 1941, Bull. Mus. Comp. Zool., 87, pp. 353-451. Revision of the African Lizards of the Family Gerrhosauridae, 1942, Bull. Mus. Comp. Zool., 89, pp. 483-543., etc. 4 bulletin: museum of comparative zoology whether the order in which the species of Cordylus are arranged, should not be largely reversed; it is difficult to believe that so well protected a species as giganteus should have given rise to less spinose forms without osteoderms, though such species might gain in activity. However, without adequate grounds for change I have accepted Boulenger's order though attracted by the idea that Platysaurus, through Pseudo- cordyhis, gave rise to some species of Cordylus like caeruleopunctatus lacking osteoderms. Cape Province, inhabited by 19 species or races, or the mountainous Transvaal with 17, would certainly appear to be the centre of speciation. This opportunity is taken of expressing my deep appreciation of the kindness shown by Dr. V. FitzSimons of the Transvaal Museum in checking over my redescriptions of five of the Cordylus described by him and inserting additional information regarding them; for amending the spelling of many place names and answering numerous questions as will be seen from the footnotes. It is with much pleasure that one of the two novelties: Platysaurus guttatus fitzsimonsi subspec. nov. is named after the author of "The Lizards of South Africa." The other: PSEUDOCORDYLUS LANGI Spec. nOV. is named for Mr. Herbert Lang whose activities have enriched so many collections and greatly contributed to our knowledge of the reptiles of Africa. I wish also to thank Dr. Walter Rose for generous permission to re- produce figs. 2 and 3, on pis. vi and xii respectively, which are taken from his book "Veld and Vlei." loveridge: African lizards Geographical Distribution of the family Cordylidae Cordylus giganteus w. warreni " w. barberlonensis " w. perkoensis " w. vandami w. depressus w. laevigaius w. breyeri caeruleopunclalus vittifer " c. rivae " c. Iropidosternum c. Jones ii " c, angolensis c. rhodesianus c. lawrenci " c. tasmani c. minor " c. cordylus c. niger peersi ukingensis macropholis cataphraclus " pusfulatus namaquensis campbelli " p. polyzonus " p. Jordani Pseudocordylus capensis robertsi langi spec. nov. m. melanonotus m. namaquensis m. microlepidotus m. fasciatus Platysaurus g. rhodesianus g. fdzsimonsi subsp. nov. g. guttalus g. lorquatus g. wilhelmi g. minor " g. orienlalis capensis Chamaesaura aenea a. anguina a. lenuior miopropus macrolepis X Total number of forms recorded c a 3 JB o £ PQ — a - 5. Si _2 Z •J = < 0 3 O 17 19 6 bulletin: museum of comparative zoology Index to the Species Recognized Page *Cordylus giganteus Smith 16 *Cordylus warreni warreni (Boulenger) 19 *Cordylus warreni barbertonensis (van Dam) 20 Cordylus warreni perkoensis (FitzSimons) 21 *Cordylus warreni vandami (FitzSimons) 23 *Cordylus warreni depressus (FitzSimons) 24 Cordylus warreni laevigatas (FitzSimons) 26 Cordylus warreni breyeri (van Dam) 27 *Cordylus caeruleopunctatus (Methuen & Hewitt) 28 *Cordylus vittifer (Reichenow) 30 Cordylus cordylus rivae (Boulenger) 32 *Cordylus cordylus tropidosternum (Cope) 33 *Cordylus cordylus jonesii (Boulenger) 36 Cordylus cordylus angolensis (Bocage) 38 *Cordylus cordylus rhodesianus (Hewitt) 40 Cordylus cordylus lawrenci (FitzSimons) 41 *Cordylus cordylus tasmani (Power) 42 Cordylus cordylus minor FitzSimons 43 *Cordylus cordylus cordylus (Linnaeus) 44 *Cordylus cordylus niger Cuvier 48 *Cordylus peersi (Hewitt) 50 *Cordylus ukingensis (Loveridge) 51 *Cordylus macropholis (Boulenger) 52 Cordylus cataphractus Boie 53 Cordylus pustulatus (Peters) 57 * Cordylus namaquensis (Methuen & Hewitt) .".... 58 Cordylus campbelli (FitzSimons) 59 Cordylus polyzonus polyzonus Smith 60 *Cordylus polyzonus jordani (Parker) 64 * Pseudocordylus capensis (Smith) 70 * Pseudocordylus robertsi (van Dam) 71 * Pseudocordylus langi spec, nov 73 * Pseudocordylus microlepidotus melanotus (Smith) 75 Pseudocordylus microlepidotus namaquensis Hewitt 78 * Pseudocordylus microlepidotus microlepidotus (Cuvier) 79 * Pseudocordylus microlepidotus fasciatus (Smith) 80 *Platysaurus guttatus rhodesianus FitzSimons 86 *Platysaurus guttatus fitzsimonsi subspec. nov 88 * Represented in the collections of the Museum of Comparative Zoology; examples of species without asterisk are earnestly desired. * LOVERIDGE: AFRICAN LIZARDS 7 *Platysaurus guttatus guttatus Smith 89 *Platysaurus guttatus torquatus Peters 91 *Platysaurus guttatus wilhelmi Hewitt 92 *Platysaurus guttatus minor FitzSimons 93 Platysaurus guttatus orientalis FitzSimons 94 *Platysaurus capensis Smith 96 *Chamaesaura aenea (Fitzinger) 101 *Chamaesaura anguina anguina (Linnaeus) 102 *Chamaesaura anguina tenuior Giinther 105 *Cha?naesaura miopropus Boulenger 107 *Chamaesaura macrolepis (Cope) 108 Family CORDYLIDAE 1845. Zonuridae Gray (part), Cat. Lizards Brit. Mus., p. 45. 1845. Chamaesauridae Gray, Cat. Lizards Brit. Mus., p. 61. 1884a. Zonuridae Boulenger, Ann. Mag. Nat. Hist. (5), 14, pp. 119, 121. 1923. Zonuridae Camp, Bull. Am. Mus. Nat. Hist., 48, pp. 297, 331. 1937b. Cordylidae Mertens, Abhand. Senckenberg. Naturf. Ges., No. 435, p. 8. For further synonymy see Boulenger, 1885e, Cat. Lizards Brit. Mus., 2, p. 251, from whom the following definition is adapted after incor- poration of most of the findings of Camp (1923), who should be con- sulted for further details. Habit robust, moderate, or serpentif orm ; head covered with sym- metrical shields; eyes present; eyelids well developed; tympanum dis- tinct or deeply sunken; dentition pleurodont, teeth numerous, small, hollow at the base, with long cylindrical shafts; palate toothless; tongue short, villose, scarcely protractile, entire or very feebly nicked at the end; body scales, if not granular, forming regular transverse series; lateral fold present or absent; limbs well developed or rudi- mentary or absent anteriorly; femoral pores present (though often in- distinct in females) ; tail stout and spinose, or moderate, or excessively long and fragile. Skull with both postorbital and frontosquamosal arches present; head with dermal bony shields in Cordylus; nasals distinct; frontal, parietal, and premaxillary single; lower jaw composed of six bones; palatines and pterygoids widely separated medially, both bordering the infraorbital fossa; postorbital and frontosquamosal arches bony; 8 bulletin: museum of comparative zoology supra temporal fossae roofed over by dermal ossifications ; zygosphenal articulation rudimentary; clavicle slender, not dilated proximally, in- tercla vicle cruciform with tendency to dilation ; sternum without fonta- nels; no abdominal ribs; non-tuberculate osteoderms present in most species of Cordylus. Range. Africa south of Giacorsa, Ethiopia, i.e. about 5° N. (Re- ported from Madagascar in error by Cope). Remarks. Camp (1923) places CORDYLIDAE (as ZONURIDAE) in a super-family Zonuroidea at the end of the section Anguimorpha, division Autarchoglossa of the suborder Sauria, thus removing it far from the IGUANIDAE, but placing it near the ANGUIDAE, the two families between which it was accorded an intermediate position by Boulenger (1885e). He would further recognize two subfamilies, i.e. Zonurinae and Chamaesaurinae, the former would now have to be termed Cordylinae. Key to the Genera 1. Habit serpen tiform; limbs short, often rudimentary, or fore limbs absent; uninjured tail at least thrice the length of head and body; ear-opening moderate or small; no collar fold; ventrals lanceolate like dorsals Chamaesaura (p. 98) Habit normal; limbs well developed, pentadactyle; tail less than twice the length of head and body; ear-opening large; a collar fold at least laterally distinct; ventrals squarish or transversely enlarged 2 2. Dorsal lepidosis entirely granular; usually a color pattern of three, light, longitudinal lines, or series of spots, on back, though sometimes absent in old males Platysaurus (p. 82) Dorsal lepidosis heterogeneous or composed of large scales 3 3. Nape covered with granules, or if dorsals extend to occiput they are much reduced; back usually covered with granules intermixed with nodular scales or dorsals, the latter small, soft, and devoid of underlying osteodermal plates Pseudocordylus (p. 66) Nape covered with well developed scales like the dorsals, the latter strong, usually1 with underlying osteodermal plates Cordylus (p. 9) 1 Absent in caeruleopunctatus. loveridge: African lizards 9 Genus Cordylus 1763. Cordylus Gronovius, Zoophyl., p. 13 (type cordylus Linnaeus). Set aside by action of International Commission of Zoological Nomencla- ture (Opinion 89). 1768. Cordylus Laurenti (part), Syn. Rept., p. 51 (type verus Laurenti = cordylus Linnaeus). 1820. Zonurus Merrem, Versuch Syst. Amphib., p. 57 (type cordylus Lin- naeus). For further synonymy (but omit Hemicordylus Smith), see Boul- enger, 1885e, Cat. Lizards Brit. Mus., 2, p. 252. Head and body depressed; limbs well developed; tail moderate. Head shields regular; nostril pierced in the nasal; eyelids well de- veloped; ear-opening large; 4 parietals; sides of neck covered with scales; no collar fold; dorsals large, usually bony1, forming regular transverse series extending to occiput; ventrals large, quadrangular or subtriangular, juxtaposed or imbricate, forming longitudinal and transverse series; femoral pores present in both sexes; digits slightly keeled inferiorly; tail spinose. Those species which I have examined appear to have the following characteristics in common, consequently these have been omitted from the specific descriptions. Head longer than broad; rostral at least twice as broad as high; postnasal present only as an aberration; a preocular; anterior supra- ocular longest, the second broadest; frontal subpentagonal or sub- hexagonal, slightly broader anteriorly; occipitals present (except in caeruleopunctatus) ; mental rather large; a slight lateral fold; a pair of enlarged preanals (though in p. polyzonus and p. jordani occasionally subequal, and said to be so in the juvenile type of c. rivae); limbs above with large, keeled spinose, imbricate scales. Range. Africa, in savanna areas south of Ethiopia, i.e. about 5° N. Remarks. Stejneger (1936b, p. 137) has set forth the reasons for recognizing Cordylus Laurenti, in contradistinction to Cordylus Gronovius, and regarding Merrem's monotypic Zonurus as a synonym. Despite the fact that for half-a-century herpetologists have been pointing out the instability in shape (pentagonal, hexagonal, etc.) of head shields in this genus, and consequently their uselessness for taxonomic purposes, others continue to stress these unimportant varia- tions as if they were of value. Which upper labial is lowest and which 1 No osteoderms in young giganteus {fide Broom) , or caeruleopunctatus which, in the absence of occipitals and reduced nuchals, shows affinities with Pseudocordylus. 10 bulletin: museum of comparative zoology highest also appears largely fortuitous and might well be dropped from future descriptions though I have included here all the available data for some species. It might be worth pointing out that the subocular descends to the lip in the warreni group, c. niger, peersi, and the two forms of polyzonus. Suboculars normally number 3, with 2 or 4 occurring as variants; in the East African forms of C. cordylus, however, 2 has become normal at the expense of the third which has been reduced and pushed to one side, 2 also appears to be normal for a dwarf form (macropholis) known only from Kleinzee, Little Namaqualand. Van Dam has recorded 5 for C. barbertonensis, but this is apparently due to his inclusion of a small scale not truly a subocular. Upper labials are normally 5 or 6, with 7 occurring as a common variant ; only in the dwarfed macropholis are they reduced to 4 with 5 still the normal. Lower labials again are normally 5 or 6, with 7 occur- ring as a rare variation, here it is the dwarf ukingensis (known to me only from the holotype) which has reduced to 4. In the hope that the results might prove useful I counted the gulars lying between the angles of the jaws in all specimens in the Museum of Comparative Zoology, however as counts were not available for eight species I was unable to take advantage of this character. The term postfrontals is employed as preferable to frontoparietals used by Boulenger and some others. As the head is longer than broad in all species except in old males of cataphractus, where it is as long as broad, this character has been dropped from the descriptions. It will be noted that in the majority of species the keels, or spines, on the dorsum and dorsal aspect of the tail are less developed than on the flanks or sides of tail, hence it has not been considered necessary to repeat this for each species. The relative smoothness of the dorsals supposedly facilitates the lizard's entry and withdrawal from beneath boulders, while the lateral spinosities, especially when the lizard in- flates, tend to catch on the sides of a rock crevice and prove a supple- mentary hindrance to the owner being withdrawn. The state of the "lateral fold," characterized as "distinct," "weak," "slight," or "strong" by various authors, is largely a reflection of the state of nourishment of the individual being described. In an emaci- ated lizard taken after a prolonged drought it will appear "strong," in a gorged specimen or gravid female, scarcely perceptible. For this reason I have ignored the lateral fold for taxonomic purposes, though undoubtedly some degrees of variation in development occur as be- tween certain species. loveridge: African lizards 11 Color in members of the genus Cordylus seems to be particularly susceptible to environmental changes, as is evidenced by Peers, notes on cataphractus (vide p. 55). Too much importance, therefore, should not be attached to color in this genus. This lizard — cataphractus — is the only member of the genus about which we have anything approaching a complete life history, and it is to Peers (1930), whose fascinating account should be consulted for further details, that we are indebted. His paper might well serve as a model to stimulate young South African naturalists to concentrate on other members of the genus with a view to filling in the deplorable gaps in our knowledge of the family. Key to the Species 1. Lower eyelid opaque; supranasals absent 2 Lower eyelid with transparent disk; supranasals present or absent (in one species only) 22 2. Rostral in contact with frontonasal; occipital spines present 3 Rostral in contact with frontonasal only rarely, usually separated from it; occipital spines absent 6 3. Median subocular not reaching lip; occipital spines enormous; scales beneath forelimbs smooth; caudal whorls subequal, gradually diminishing towards tip of tail; range: Transvaal; Orange Free State; Cape Province giganteus (p. 16) Median subocular reaches lip between two labials; occipital spines short; scales beneath forelimbs keeled; large caudal whorls separated by smaller ones 4 4. 34-42 dorsal scales between occiput and base of tail; range: Zululand w. warreni (p. 19) 24-32 dorsal scales between occiput and base of tail; range : Transvaal 5 5. 14-16 ventral scales across the belly races of warreni1 10-12 ventral scales across the belly; range: Waterberg Mountains, northwestern Transvaal w. breyeri (p. 27) 1 See full descriptions of various races found in mountains of eastern Transvaal. 12 bulletin: museum of comparative zoology 6. No occipitals; nuchal scales of first six rows greatly- reduced, minute; range: Cape Province caeruleopunctatus (p. 28) Occipitals present; nuchal scales moderate or large. . . 7 7. Nuchals comprising foremost, i.e. postoccipital, row, twice as large as those in second row; range: Bechua- naland; Transvaal; Zululand; Natal vittifer (p. 30) Nuchals comprising foremost, i.e. postoccipital, row, subequal, certainly not twice as large as those in second row 8 8. Interparietal enclosed between two pairs of parietals . . 9 Interparietal on a line with anterior parietals and in contact with (rarely separated from1) the postfrontals . 18 9. Dorsals elongate, the two vertebral rows not or but scarcely enlarged; laterals unlike dorsals 10 Dorsals squarish, the two vertebral rows much en- larged; laterals like dorsals 15 10. Head moderately depressed; temporals moderate; gulars small 11 Head much depressed; temporals large; gulars moder- ate or large •. 14 11. Femoral pores 3; nostril said to be pierced in centre of nasal; range: Ethiopia c. rivae (p. 32) Femoral pores 5-8; nostril in lower centre of postero- inferior corner of nasal; range: Kenya Colony and countries to the south of Kenya 12 12. Ventrals in 28-34 transverse rows; size larger, length from snout to vent in adults 82-92 mm. ; range : Kenya south to Mozambique and Southern Rhodesia c. tropidosternum (p. 33) Ventrals in 22-27 transverse rows; size smaller, length from snout to vent in adults 64-74 mm 13 13. Southern Mozambique; Southern Rhodesia (south of Bulawayo); Bechuanaland Protectorate (also intro- duced at Kimberly, Cape Province) c. jonesii (p. 36) Southwestern Angola (known only from scanty description); probably northern South West Africa; possibly north-central Belgian Congo also c. angolensis (p. 38) 1 Sometimes in namaquensis. loveridge: African lizards 13 14. Posterior parietals much larger than the anterior; dorsals in 27-28 transverse rows; ventrals in 14 longi- tudinal rows; range: Southern Rhodesia northeast of Bulawayo c. rhodesianus (p. 40) Posterior parietals subequal to the anterior; dorsals in 24 transverse rows; ventrals in 12 longitudinal rows; range : Little Namaqualand c. lawrenci (P- 41) 15. Laterals on flanks slightly smaller than dorsals; range: Uitenhage Division, Cape Province c. tasmani (p. 42) Laterals on flanks just as large as the dorsals 16 1 6. Rostral 2 3^-4 times as broad as high ; median subocular not, or but rarely, descending to the lip; color above brown or olive with or without markings; range: Cape Province (exclusive of parts of Uitenhage Division) . . c. cordylus1 (p. 44) Rostral, 2-2 x/i times as broad as high; median sub- ocular descending to the lip between fourth and fifth labials; color above uniformly jet black 17 17. Head shields smooth or slightly rugose; nasal moderate, not or but slightly swollen; temporals large, rarely keeled; the two vertebral rows of dorsals enlarged; femoral pores 5-9 ; range : Cape Peninsula c. niger (p. 48) Head shields very strongly rugose; nasal large, very strongly swollen; temporals very large, strongly keeled; vertebral rows of dorsals not differentiated from adjacent dorsals; femoral pores 8-12; range: Little Namaqualand peersi (p. 50) 18. 11-14 gulars between angles of jaws; the two vertebral rows of dorsals enlarged; 14-18 longitudinal rows of dorsolaterals; 9-12 longitudinal rows of ventrals 19 16-21 gulars between angles of jaws; vertebral rows of dorsals not enlarged; 20-32 longitudinal rows of dorsolaterals; 14-29 longitudinal rows of ventrals. ... 20 1 See also the recently described C. c. minor from Matjesfontein, C. P., said to differ from c. cordylus in having 24-36 longitudinal rows of dorsals, 16 rows of ventrals, only 4-6 femoral pores, and smaller size. 14 bulletin: museum of comparative zoology 19. Gulars small, the anterior not enlarged; dorsals in 30 transverse rows; ventrals obtusely keeled, in 30 trans- verse rows; range: Southern Tanganyika Territory. . .• ukingensis (p. 51) Gulars very large, a few anterior ones moderately en- larged followed by a zone of much smaller ones; dor- sals in 16-19 transverse rows; ventrals smooth in, 20-22 transverse rows; range: Little Namaqualand. . . macropholis (p. 52) 20. Dorsolaterals in 20-22 longitudinal rows and 15-16 transverse rows; range: Little Namaqualand cataphractus (p. 53) Dorsolaterals in 24-32 longitudinal rows and 30-32 transverse rows 21 21. Ventrals in 14 longitudinal rows; range: Central South West Africa pustulatus (p. 57) Ventrals in 18 longitudinal rows; range: Southern South West Africa namaquensis tp- 58) 22. Supranasals absent; range: South West Africa campbelli (p. 59) Supranasals present; range: South West Africa 23 23. 10-19 femoral pores; upper posterior femorals scarcely larger or more spinose than the anterior; whorls of large scales at middle of tail separated by whorls of smaller ones; a dark streak on side of neck; range: South West Africa south of Aus p. polyzonus (p. 60) 5-8 femoral pores; upper posterior femorals much larger and more strongly spinose than the anterior; whorls of large scales at middle of tail diminishing gradually towards tip of tail ; no dark streak on side of neck; range: South West Africa north of Aus p. jordani (p. 64) loveridge: African lizards 15 sajod jBJocaa j CM CM ** CM in -, — , SO _ C\ ,-*,_i,-if-i__,., rHeo.oecr-^1- t-\DCTvas"t~-2^ 2 'He0 — 1 1 1 1 2 2 2 1 L"l 1 1 1 co l l 1 7 1 7 1 c~ 7 ■* 1 L HooMoHHH«\6 uiuiinm *>j«uiJ,n,l,N A © ** sjBJinaA jo smoj gsjaASUBjj^ CO in SO in s© tJ" CM © so C\ CJ\ Cs SO © Os O NOtnOCSOO MMM_«0„MMninMMININnMINnM(N0INniNm«'f* inco-i^i>MM4'COvo^eo^MM"co4f!:J5M" | 1 ~ 1 1 I 1 von cmso© IN^COmC0^0;fATt'CM-»CM | msosom in in in in in sjBjnooqng I CO CO CO CO CO CO CO CO CO CM CM CM CO CO CO CO CO CO CO CO CO CO ICOCOCO ICO pm CO CO saijBij'OBjdng m># mm ■# *p tj *ji COI I't'^'inl |CO|CO|COC"COCOCOCOCOCO|COICOC"COCOCOCO COCO Tf Tf CO CO CO CO sjBpnooBjdng CO ^qSiaq oiui qjpsajq |BJjsoy V*\jJ" \W \W SHI \rf| w\oi\ ,4\ -*\ \n CO CO ■<* V* «\\w NHI «P?-«\ CO CM CM CM \jf\TlCMTji -JVCO 1 Vf 1 1 «xV»CM-\ -\\j» 1 CO 1 1 1 CM 1 CM -*s-^ 1 1 *-• CM 1 C" \j).-\vff)C-. \«CM CM -\ 1 CM c- CM -\\* 1 CMVJXCl Vt CMCMVfCO 1 CM -J\CM-<\ p<\ 1 CM Vf I 1 CM -^\f< Js_K J\ -K IN CM CM CM CM -\M CM CM ->\ (N CI CM