Frictional Fire-Making with a Flexible Sawing-Thong

Frictional Fire-Making with a Flexible Sawing-Thong. Author(s): Henry Balfour Source: The Journal of the Royal Anthropological Institute of Great Britain and Ireland, Vol. 44 (Jan. - Jun., 1914), pp. 32-64 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/2843530 . Accessed: 07/01/2015 14:55

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend access to The Journal of the Royal Anthropological Institute of Great Britain and Ireland.

http://www.jstor.org

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 32

FRICTIONAL FIRE-MAKING WITH A FLEXIBLE SAWING-THONG.

BY HENRY BALFOUR.

AMONGthe variousmethods whereby fire is obtainedby frictionof wood, there is one whose varietiesand geographicaldistribution have not hithertobeen described with any degree of comnpleteness,although its ethnological significanceis an importantone. I have, therefore,thought that it mightprove of use if I were to bringtogether rny notes on the subject,with a view to placing on record the wide range of this special method,as faras I have been able to trace it througha study of the literatuireand of specimenspreserved in inuseums and private collections. I hope that the descriptionsand the maps which I offermay be the means of elicitingfurther data, and that thus the complete geographicalrange of this fire- makingappliance may becomeknown. The better-knownfrictional fire-making apparatus which involves the employ- ment of a rigid,blade-like saw of bamboo or wood, is to be traced over an area which,to put it briefly,extends from India and Assam, though the Indo-Chinese region,the Malay Peninsula,and the Asiatic Archipelago,to Australia. It is also employedin the Nicobar Islands and in NorthAmerica, and has been claimed for one districtin CeintralAfrica. In Europe,too, it has been used in variousregions. I do not proposehere to give a detailed descriptionof its geographicaldispersal, buitmerely refer to it as beinga sawingmethod which is closely allied to the one underconsideration, and whosegeographical r'ange coincides to a verygreat extent with that of the latter. In the East, at any rate, it seems probable that the flexiblesaw, as applied to fire-making,is merelyan ilnterestingvariant of the rigid saw, and that these twomethods have one comrnonorigin. The essential differencebetween these two " sawing'"methods is to be seen in the "saw " itself,which in theone is a rigidblade or stick, and in the other is a flexibleband or thongu.In nearly all cases where the flexiblesaw is employed, the procedureis essentiallythe same. A stoutpiece of bamboo or wood is held in position,and a narrowthong of cane or othersuitable material is looped roundit, the frictionbeino caused by pulling the ends of the thongalternately, so that it is sawn across the stick or post, detachingi,in the process a fine wood-powderand generatingrby the frictionsufficient heat to ignite the powder and cause it to smoulder. From the smoulderingdust a flame can readily be obtained with the help of some quicklyinflamtmable material, such as dry grass,etc. The details and accessoriesof the processare referredto in the various local descriptionswhich I

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions F'rictiona -Fire-makingwith a FlexibleSawing-thong. 33 quote below in the order of their geographicaldispersal, which, as faras I have hithertoascertained, is as follows:- NAGAHILLS, ASSAM.-The north-westerlyextremity of the main area of geographicalrange of the practice of fire-makingwith the flexiblesaw is to be foundamongst the NAGAS of the Naga Hills, who adopt this method,at any rate locally. Referenceto this regionwas given by the late Mr. S. E. Peal, formerly residentin Sibsagar,who wrote' that firewas obtainedby the Nagas (or Nogas, as he preferredto call them) " by means of a long piece of cane passed under a dry log and pulled (i.e.,the ends) alternatelyby the rightand left hand,so as to ignite somietinder placed in a hollow or split underneath"(see Fig. 1).

NAGA, ASsAM. 1 FIG.

a ~~~~~~~~~~~~~~~~~~~~~~~Ci NAGA, ASSAM NA A,SSAM. FIG. 2. FIG. 2A.

In the EthnologicalMuseum at Berlin I saw a Naga fire-makingapparatus of this kind (Fig. 2), the native names forthe various parts being given as follows, yung= thecane strip,mniset = the woodenstick, renr-a = the tinder. I have also just received,through the kindnessof Mr. J. H. Hutton,a complete fire-makingset fromthe Naga Hills, consisting(Fig. 2A) of a billet of lime wood, split at one end and wedgedopen with a stone. The thongis a narrowsliver split froma piece of bambooand whittledsmooth, the shavingsfrom this supplyingthe tinder,which is placed under the fork. The billet is held down on the ground

1 Joturn.Antkrop. lnst., xxii, 1893,p. 252, and Plate XVIII. VOL. XLIV. D

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 34 HENRY BALFOUR.-FrictionalFire-makcing

witlhthe foot,and the thoingis sawn roundit in the usual way. Aliuost any wood may be used but lime is preferred. CHITTAGONGHILLS.-Captain T. H. Lewin' describesin detail this metlhodas practisedby the TIPPEKAH tribeof theChittagong hill-tracts. "They take a piece of drybamboo about a footlong, split it in half,and on its outerround suirface cut a nick or notch,about Sth inch broad,circling round the semi-circuinferenceofthe bamboo, shallow towardsthe edges,but, deepening in the centre,until a miniuteslit about a line in breadthpierces the innersurface of the bamboofire-stick. Then a flexible stripof bamboo is taken,about 12 feetlong aind 1th inch in breadth,to fit the circlinignotch or groovein the fire-stick. This slip or band is rubbedwith fine, dry sand, and then passed roundthe fire-stick,on whichthe operatorstands, a foot on eitherend. Then the slip, graspedfirmly, an end inl each hand, is pulled steadily back and fortb,increasing gradually in pressureand velocity as the smoke comes. By the time the fire-bandsnaps with the friction,there ought to appear through the slit in the fire-sticksome incandescentdust, and this,placed smoulderingas it is in a nest of drybamboo shavings, can be gentlyblown into a flame." ANNAM.-Awayto the eastward,the same processof makingfire is adoptedby the o4ois,a primitivepeople inhabitingthe table-landand mountainsbetweenl the Mekong River and the Annam coast,from the fronatiersof Yunnan to Cochin-China. They are probablyof Indonesianstock. The processis almost identical with that followedby the Tipperahs. Accordingto M. A. Gauthier,2half of a split bamboo is notchedon the convex surfaceand is held downlby the feet. Strips of green bamboo-rindare split off,c. 8 mm.-thick and 50 cm. long. To each end a little woodentoggle is fixed. About 100 of these stripsare usually prepared at a time. Fire is obtainedonly after expending some 30-40 stripswhich break afterseven or eightsawing movements. The methodis not common,as slow-matchesare usually carried,and the Mois also employthe fire-piston.3 MALAY PENINSULA.-In thisregion, the SAKAI, SEMANG and MALAYSall employ this method. W. Skeat and C. Blagden4 mention that it is a common method among both the Western,or Kedah Semang,and the Eastern Semnang,or Pangan, who use a rattanthong and a driedbranch (of certainspecial treesonly) which is held down with the feet. T. N. Annandale and H. C. Robiinson5give a full descriptionof the process as witnessedby themin Grit in the Upper Perak district,where the older men still resort to it in place of more modern methods. They describe how a SEMANG chief" took a billet of softwood, about 1I feetlong, and split it at one end so as to forma cleft of about 6 inches. Into this he inserted a small stick, which

The WVildRaces of S.E. India, 1870, pp. 207, 208. 2 Bull. Soc. Geog. du Havre, xiv, 1901, p. 95. v. H. Balfour, " The Fire-piston," in AnthropologicalEssays presented to Edward Burnett Iylor, Oxford, 1907, p. 27. 4 Pagan Races of the Mal7ayPeninsula, 1906, i, p. 114. a "$Fasciculi Malayenses," Anthrop.,pt. i, 1903, p. 15.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawving-thong. 35 formeda peg separating the two halves and standing,above the surface of the billet to the heightof an inch or more. Beside this he placed some 'palmn-scurf.' He then took a stoutstrip of rattan, about 5 feet long,and passed one end of it under the billet as it lay on the ground. To each end he fastened a stick, which acted as a handle (of. the Mois above). Then he grasped one of these sticks in each hand,and, holdingdown the cleftbillet by means of his rightfoot, he began to draw the rattanbackwards and forwardsacross the inner surfaceof the billet. He graspedthe peg whichkept the cleft open betweenthe great and second toe. The frictioncaused by the rattanrubbing against the soft wood soon produced a, considerableamiount of heat, whichfirst blackened the wood and then caused the tinderto take fire. Liftingup the billet,the man had no difficuiltyin lightiiiga cigaretteat the palm-scurf,which was now smoulderingin the cleft." A specimen of this apparatusfrom Upper Perak is shownin Fig. 3.

FIG. 3.

In theBritish Museum there is a fire-makingset fromthe Semangof Grit,with a sawing-thongof bamboostrip, the ends of which are fittedwith small wooden toggles(Fig. 3A), as describedabove in connectionwith the Semang fire-sawalnd also with that of the Mois.'

MNALAY STA-TES FIG. 3A.

The same authors2describe a similarprocess amongthe MAI DARA'T(SAKAI) of the Batang Padang districtof SouthernPerak. The actual procedurediffered slightlyfrom that seen among the Semnang,as the woodenbillet was not split and no peg was inserted. The tinder was held near the grooveformed by the rattan thong. "The ends of the rattanwere held by a man sittingon the ground,and the same man shovedagainst the billet of wood with his riglhtfoot, thus keeping it pressedhard against the rattan,which he drewbackwards and forwardsround it.

1 Handbookto theEthn-ographical Collections, 1910, Fig. 7c. 2 "Fasciculi Malayenses,"Anthrop., pt. i, 1903,p. 14. D 2

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 36 HENRY BALFOUR.-FrictionalFire-making

This method is consideredmost efficientif suitable materialscould be obtained; but verystrong rattan was necessary,as well as peculiarlysoft wood." Alternative methodswith the Mai Darat are fire-making,with the stiffbamboo saw or with the thong-drill. A specimenfrom the Mai Darat is in the BritishMuseum. Captain G. B. Cerrutilgives a somewhatdifferent version of the processas seen by him amongthe SAKAI. " A bambooreed is taken and a small holeis made in it, into whichis introducedthe tinder,a downysubstance foundat the base of palm-leaves and known by the name of lulup amongthe Malays. Round the bamboo is wound two or three times a long stripof veryflexible rattan. The operatorholds in his hands the two ends of the rattan,and, pressingfirmly upon the -bamboowith his feet, commencesto pull firstone end and then the other, forciblyand with great rapidity" (Fig. 4). Leaves of dry grass are thrown on to the tinder when this has ,(, >) become ignited. The way in which the smoke of the resultantfire behaves is observed and is used fordivininig whethera selectedsite fora new habitation is auspicious or not. I must confessto feelingsome doubt as to the cane thong being " wound two or three times" roundthe bamboo, since it is likely that the frictionwould be too great to admit of rapid sawing,and to pass the cane thong more than once roundis contraryto the practice of other peoples who employthis method. _ __ - The MALAYS of the Peninsula are mentionedby S AKAI. Nelson Annandale and by Leonard Wray as employing FIG. 4. at times the flexible-saw apparatus for fire-making, though ordinarilyother imoremodern methods are available. A specimenof the Malayan apparatus(Fig. 5) was sent to me in 1892 by Mr. Leonard Wray,and is now in the Pitt Rivers Museum. It is practicallysimilar to the Semang apparatus

M1PLAY, PERAK. FIG. 5.

(Fig. 3) and consists of a stout stick of very soft,white wood, split at one end and with a woodenplug insertedin the cleft,to keep the two halves apart,a collar of rattan preventingthe slit from extending. Tinder formedof bamboo shreds is insertedin the cleftat the point wherethe sawing takes place. The saw is a long

1 NVelPaese dei T'eleni,1906, p. 161; and in the translatedversion, My Friendsthe Savages, 1908,p. 146.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 37 and flexiblecane thong. In using the apparatus,itheMalay sets the stickvertically against a tree,fixing the, upper enidby bindinga cane strip roundit and the tree (Fig. 6). The lower end is steadied with the foot,and the ends of the cane thong which is passed round the stick at the spot where the tinderis, are alternately pulled horizontally. BORNEO.-Passing eastward, as I have no reference to the occurrenceof this method in Sumatra, the next reaion to be noted is Sarawak and the adjacent parts of British North Borneo and Central Borneo,where it is employed nowadays ceremoniallyby the KAYANS and KENYAHS, especially by the former. W. H. Furness' describesthe ceremoniesobserved in connectionwith the naming of a child at the house of Tama Bulan, a Kenyah chief,living about 400 miles inland on the Pata River,a tributaryof the Baram. New-firehad to be procured by friction. He says,"The Musa consistsof a piece of softfibrous wood, which is

KENYAH

MALAY, PERAK

FIG. 6. 7. held down by the feet,firinly on the ground,and restsupon a bundle of fineslivers of drywood; underneathit is passed a stripof drybut flexiblebamboo, whichis sawed back and forth until the frictionstarts a spark in the finedust whichhas been therebyrubbed up; the spark is fosteredand soon blowninto a flamein the bundle of slivers. When the materials are in proper condition,fire can be produced in much less than a minute. Should all the firesin a house go out,or when fireis to be started forthe firsttime in a new house,the Musa is the only method wherebyfire may be kindled,-no flintand steel, nor fire-drill,nor fire- syringe,nor matches(common enough, thanks to the Chinese bazaar), can be used; it mustbe the Musa, and the Musa alone. At the namingof a child, the piece of soft wood is carved into a grotesque head at one end. The image thus nmadeis called 'Laki Pesong,' the god of the Musa." In Fig. 7 I reproducethis author's

1 HomeLife ofthe Borneo Head-hunters, 1902, p. 37.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 38 HENRY BALFOUR.-FrictionalFire-making figureof the apparatus. It will be noted that in this example,the stickis not split at oile end, but a narrowlongitudinal slit is cut throughit and answersthe same purposeas a cleft. A numberof the bamboo stripsare shown. Mr. Furness also refers to the Musa in Folk Lore in Borneo' and mentions that it is also employed when consultingthe omen birds. It appears that divination is also practisedin conjunctionwith the productionof lnew-fireat a child-naming,since, if the cane saw, at the momentwhen the spark is obtained,breaks in even lengths, the omen is regardedas unluckyand a new name must be chosen,whereas if the two pieces are of unequal size, the omen is auspicious alnd the name is a good one.

FIG. 8. FIG. 9.

Fig. 8, taken froma photographgiven me by the late R. Shelford,shows a Kayan makingfire in the mannerdescribed. Figs. 9, 10 and 11 are fromspecimens collectedby Mr. C. Hose in 1904 and 1906 fromthe Kayans of the Barain district, Sarawak,and were made foruse duringthe child-namingceremony. One specimen (Fig. 9) is completelyanthropomorphic. Fire has been obtained by sawingthe

FIG. 10. cane thong round one of the legs of the figure, and not across the slot which separates the legs. Mr. Hose states that the figurerepresents Laki IPesong,the god of fire,to whom Kayans pray for long life for a child,and he adds that the

1 1899,pp. 8, 10 and 24.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 39 carved wooden piece is called pesong,the cane thongpusa. The second specimen (Fig. 10) is also anthropomorphic,the head of Laki Pesong being carved at each end, the longitudinalslot at the centrerepresenting in a conventionalmanner the fusedlegs of the two figures. The third example (Fig. 11) is not carved,but one end is split,the cleftbeing wedgedopen with a piece of wood. As in the example shown in Fig. 9, the firegroove appears on one limb only of the fork,and does not extend across the cleft. Some palni-scurftinder accompanies this specimen. A similar apparatus is in the BritishMuseum. Hose and McDougalll describeand figureanother example of the Kayan apparatus(Fig. 12) in which Laki Pesono

KAYAN KAY A14 FIG. 11. FIG. 12. is representedas a squatting figure,carved at one end. "For each child who is to be named a small humanimage in softwood is prepared. This is an effigyof Laki Pesong, the god whose special functionit is to care forthe welfareof the children. A small mat [shownin Fig. 12 and also in Fig. 9] is wovenand a few stripsof rattanprovided for each child. Each child sits with his (or her) mother in the gallerybeside the door of theirroom, and the parents announce the name they propose for the child. Then the father,or some other man, after killing a chick or younig pig, lays the image on the mat before the child, passingone of the rattanstrips beneath it, and, holdingthe image firmlywith a big toe on each end of it, pulls the strip rapidlyto and fro,until it is made hot by its frictionagainst the imiiage,and smoke begins to rise." The cane thongbreaks eventually and the auspiciousnessor inauspiciousniessof the suggestedname is divined according to whetherthe two pieces are of unequal or of equal length. If the latteroccurs, another name is suggestedaind is submittedto the same test. The name pusa appears to be applied to the whole apparatus and also to the ceremony. 1 The Pagan Tribesof Borneo, 1912, ii, p. 160,and Plate 168,Fig. 2.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 40 HENRY. BALFOUR.-FrictionalFire-making

Dr. A. W. Nieuwenhuis' figuresa specinlen obtained fromthe Kayans of Ceiltral(Dutch) Borneo,a sketchfrom his photographbeing reproducedin Fig. 13. This specimenis almost identical with that shown in Fig. 11, the stick,however, beingmerely split for half its length,the "limbs" retainingtheir rouch edges, instead of beingrounded off as in the latterexample. Also, if we can judge fromDr. Nieuwenhuis'photograph,the " saw " passes roundboth limbs, instead of round olle only,as in the morespecialized, anthropo- X /. morphicapparatus of the Kayans. Since in the other regions where this method of fire-production obtains, the band almost invariablypasses round both "limbs" of the stick, when this is split and wedged open, we nmayregard this example as re- FIG. 13. presentingthe more primitivetype in Borneo,and it seems likely that the anthropomorphicassociation in thisisland was firstsuggested by the bifurcatioinof the stick, which simulated body and legs, the human resemlblancebeing improvedupon subsequently. This involved iincidentallya shaping and roundingof the legs of the figure,and, the rough edges having been eliminated,it was probably found that the requisite friction was more readily producedby sawinground one of the rounded limbs than by sawing round both,and also that the little pile of friction-dustbecame concentratedupon smallerspace, enabling it to accumulate the heat quicker. We may,I think,account in this manimerfor what appears to be a departurefrom the normal procedure. This method appears nowadays to be restrictedin Borneo to purely ceremonial.use. For domestic purposes alternative methods of fire-production, associated withvarious tribes, are the rigidfire-saw, the hand fire-drill,percussioni with bamboo and crock,flint-and-steel, the fire-piston,and, recently,imported matches. PHILIPPINE ISLANDS.-In the Philippine group the thong-sawis again met withamong the Negritosof Casiguran,N.-E. Luzoni,whose apparatus was collected by Dr. Semperand has been described and figuredby Dr. A. B. Meyer.2 It consists(Fig. 14) of a stick,32 cm. lonig,split fortwo-thirds of its length,the two parts being wedgedapart with a piece of wood,some white bark-clothserving as tinder,and a long stripof rattan,beJuco, about a metrelong, serving as a saw. The bark-clothis fixedin the cleftof the stick and the rattanthong a is sawn roundboth liinbs of the fork in the usual manner. The specimens are in the Dresden Museum. The Mangyansof the Bako districtof Mindoro Island (which lies between

1 In CentralBorneo, 1900, i, Plate XIX. 2 Die Philippinien,publication of the Royal Museum,Dresden, No. ix, 1893, Plate II, Fig. 7, and p. 5.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a FlexibleSawing-thong. 41

Luzon and Palawan), also make firewith the rattan thong-saw. Mr. M. L. Miller writes of themas follows: "In the Bako region practicallyevery man met with, and some womenas well, wears on the leftupper arm 1 or 2 or 3 armletsof -rattan. These appear ornamental,but are used in makinigfire. A piece of dry wood about 3 cm. in diameteris selected, one end of it is split for a distance of 12 or 14 cm. and a plug is put in to keep the slit open. A few shavings are then put tightlyinto the pl slit where it is narrow. One of the armlets is taken from the arm, unwound, and passed round the split stick just under the shavings. The split stick is held firmlyon the ground with the two feet,and the rattan is drawn rapidly backward and forward until the shavings ignite. A few seconds are sufficient." Mr. Miller has traced Negritic affinitiesin this districtof Mindoro,Iand it is probable C L ~~~~~~~CASI U RAN ,L UZON . that there formerlyexisted a Negrito FG 14 populationin this island. A photographby Mrs. F. Cooper Cole showinga " pygmy" of the Island of Palawan making fire,is reproducedin The,eCustoms of the World.' A sketch fromthis photographis given in Fig. 15. The stick in this instance does not appear to be split. Althoughdescribed as beinga " pyginy," it is not certain that the native representedis a NegTito, , though it is probable that he is at any rate partially Negritic. The Island of Palawan forms a geographical link betweenN.-E. Borneo and Luzon. NEW GUINEA.-We next meet with this apparatus in the Island of New GuLinea,where its distributioncovers a wide area, which extelnds into the Dutch, British and German territories. In Dutch New Guinea the most westerlydistrict from which I have a record of the use of the flexible-sawis that of the upper watersof the Mimika River, where it was

observedby the membersof the recentexpedition organized PALAWAN ,PHILIPPINES by the BritishOrnithologists' Union, 1910-1911, among the FIG. 15.

1 1913,p. 657.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 42 HENRY BALFouR.-FrictionalFire-making

TAPIROpygmies of the lowerslopes of the mountainrange, c. 40 13' S., 1360 43' E. Mr. A. F. R. Wollaston has given a full descriptionof the process.' He also verykindly gave me two sets of the apparatus,and a thirdset to the Pitt Rivers Museum. Two of these are seen in Figs. 16 and 17. As will be seen,the stick,

TAPIRO. TAPIRO,1DUTCrH N GuNEA FIG. 16. FIG. 17. whichis quite short in all the examples,is as usual split,the clefthaving a small stone inserted so as to keep it open. The example shown in Fig. 16 is split for about three-quartersof its length; the other(Fig. 17) is split rightthrough, but the lower end is kept closed by means of a binding-collar. The thong-sawis a very long strip of rattan,which is kept neatlycoiled up in a thick ring. The tinder is usually a piece of the fibroussheath of a palm-shoot,or a piece of dry moss. The tinderis placed in the narrowpart of the cleft,behind the stone. The operatorplaces the stick on the ground,holding it down with one foot. " Then, having unwound about a yard of the rattan,he holds the coil in one hand and the freeend in the other,and looping the middle of it underneaththe stick at the pointwhere the tinder is placed, he proceedsto saw it backwardsand forwards with extreme rapidity(see Ficr.18). In a short space of time, varying from ten to thirty seconds, the rattan snaps and he picks up the stick with the tinder, which has probably / by this time began to smoulder and blows it into a flame." The thong passes round both limbs of the forked stick. 44@ge Both the specimeis figuredhave been much used and show numerous charred grooves,each of which indicates where a < jj " sawin " has been performed. The fire-sawis most frequently

-TAP RO used for lightina tobacco. The Tapiro showed no signs of FIG. 18. surprise at the Englishmen's matches, and did not wish to

1 Pyqmiesand Papuans, 1912,pp. 200, 201, 202.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 43 possess them,preferring their own primitivecontrivance, which they use with such skill and rapidity. Captain C. G. Rawling,' anotherulember of the expedition,also refersto fire- makingamong the Tapiro,and gives threeillustrations of the process. Mr. Wollast,on,during the morerecent and verysuccessful expedition to Dutch New Guineain 1913,again observedthis methodamiong the hill people,of supposed Negriticstock, on the slopes of Mount Carstensz,c. 1370 12' E., 40 10' S., livingat an elevationof between4,000 and 6,000 feet. I am much indebted to him for givino me particuilarsin a letter. The practice in this district,which lies somewhatto the east of the regionoccupied by the Tapiro,is almostidentical with that of the lattertribe, only "whereas the Tapiro carrythe thong in a long,strip twisted into a ring,the Carstenszpeople carryit, cut into shortlengths, each oile sufficientfor one operationionly. They are carriedfolded in U-shape,tied together in a bundle of a dozenor so . . . They secure the stick with bothfeet, while the Tapiro hold it down with onefoot only, but this may be only a chance This methodis not used by the people of the coast."' It is importantto note that fire-makingwith a pliant canie-sawis restrictedin Southern Dutch New Guinea to the hill people,who are eithertrue pygmies,or at any rate are largelyof Negritic origin. The Papuans of the Mimika River arid Parimau district,as also those of the coastal plains fturthereastward, do not employ this methodof procedure. Furthereastward, againi, a recordof this fire-miiakingmethod is furnishedby Dr. H. A. Lorentz,3who, during his successfulattack upon Wilbelmina Peak, 1907-1909, observedit amongthe hill nativesof the Treub Mountains,on the head- watersof the LorentzRiver (Noord River),c. 40 24' S., 1380 42' E. He does not describethe process,but merelysays " we learnthow by frictionof rotan one can make fire." While Dr. Lorentz states that he did not meet any pygmies,he differentiatesbetween the mouintainPapuans and those of the plains, saying that the former" are not so large as the people of the Lorentz(north) River." Thus, it is very possible that there muaybe traceable some Negritic element in the highlands,which would suggestan affinitybetween these people and the Tapiro, whose methodof fire-produictionis the same. On the southern slopes of Mount Goliath,40 43' S., 1390 51' E., drained by the Eilanden River, Army-SurgeonA. C. de Kock saw natives of Negritic affinity who procured fire after the manner of the Tapiro. Two pieces of'wood, about 1o cm. long,were bound togetherso as to enclose a wedge-shapedspace at one end, into which tinder fromthe leaf-sheathof Caryota.Rrnuphiana was inserted. A slender strip of rattan was passed under the forked wood and drawn to and fro.4

1 TheLand of the.Vew GuineaPygmies, 1913, pp. 111 and 273. 2 See also The GeographicalJournal, xliii, 1914,p. 262. 3 The GeographicalJournal, xxxvii, 1911, p. 489. 4Van dem Broekin Zeit.f. Ethnologie,1913, p. 28.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 44 HENRY BALFOUR.-Fr-ictionalFire-mnakcing

Continuingeastward, the boundary-linebetween Dutch and British New Guinea muistbe crossedbefore we again meet with the flexiblefire-saw. It was recordedby L. M. D'Albertis1in 1876, on the head-watersof the Fly River,above the junction with the Alice River, and near the lower slopes of the Victor Emanuel Range. He says, " I observed some round logs of wood,probably- the trunks of small trees-which bore spiral marks of fire. On examiiinationI ascertained that with these the natives are accustomedto kindle their fires. A piece of rattanabout two feetlong, and split in two forabout half way its length, used like a rope rounda drill,causes the wood to take fire by its friction." There evidentlyhas been some slightconfusion in writingup his field-niotes,and we may assume that the words" split in two forabout half way its length,"applied by him to the cane thong,should rather refer to the log. The partial splittingof the thong seems to be unnecessary,whereas, as has been seen, it is a very usual procedurein otherregions in preparingthe log or stick forthis puirpose. Moreover, D'Albertis himselffigures2 a fire-stick(though, unfortuinately, without any data in the text) which is clearly split at one end, and exhibitstwo transversegrooves across the cleft,evidently produced by sawing-friction(Fig. 19).

UPPER FLY R,NEW &=UI NEA FIG. 19.

At the mouth of the Fly River this apparatus was observed by the Rev. J. Chalmersin the Island of KIWAI.3 He gives no details,merely saying that fireis sometimesobtaiined " by a lengthof cane drawnswiftly across a piece of wood held down by the foot: the ends of the cane are held in both hanldsas among the Koiari at the back of Port Moresby." An alternativemethod of fire-making in Kiwai is the "ploughing" method. The Kiwaians appear to be true Papuails, thouch the statementsas regardstheir staturevary somewhat,and it is possible that some may show a slightadmixture of Neuriticblood. A little furthereastward and again inland in the mountainlregion, the Hon. M. StaniforthSmith discovered the flexiblefire-saw among the Bush-natives of the SAMBREGI tribes,a little to the N.W. of Mount MURRAY, liviIng at an altitudeof about 6,000feet above sea level, c. 143' 50' E., 6040' S.4 'Their niethod of making fire is superiorto the usual Papuan system. They get a piece of dry soft wood, split one end and insert a piece of tapa cloth,thenl taking a piece of cane, whichthey carry twisted round theirwaists, they place it under the wood on

1 New Guizea, 1880, ii, p. 99. 2 Loc. cit.,p. 378, Fig. 2. 3 Journ.Anthrop. Soc., xxxiii, 1903, p. 118. 4 GeographicalJournal, xxxix, 1912, p. 327, and Annual Report,British Nrew Guinea, 1911, p. 169.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 45 whichthey stand. Graspingeach an end of the cane, theypull it backwards anld forwardsvigorously; when it has eaten half way throughthe wood to the tapa cloth the heat generatedis so great that the cloth smouldersand is blown into a flame. The whole process is accomplished in tenl or fifteenseconds. I am informedthat certainniatives on the main range about Kagi adopt this systemalso; if so, it is interestingas possibly showing soine connectionbetween them." He describes the physique of the Sambregi people as magnificent,and it is unlikely that there is any marked inftusionof Negritoblood among them, although the physique is described as not being uniform. It is suggestedthat theremay be some minglingof Papuans and Paptio-Melanesiansin this district. The next locality of which I have recordis the districtof the Purari Delta, wherethe MAIPUAtribe is of Papuan stock. I have two sets of the apparatus,karo karo,from this tribein my collection,and one whichis shownin Fig. 20 is in the

MAPL!'A PURARP DELA ,NEW CGUINEA. FIG. 20.

Pitt Rivers Museun. I puirchasedall threetogether in 1899. Each set consists of one or two long and fairly stout staves or fire-sticksof very light wood resemblingHibiscus, and a bundle of stripsof cane in fairlylong lengths; two of the sets have in additiona coil or two coils of cane-strip. Also each set has two long woodenpegs, roughly pointed at onieend, as though for fixingin the ground to hold down one end of the fire-stick. The fire-sticksare not split, but are- pointedat one end. No details as to the precise methods pursued in using the apparatusaccompanied the specimens,but we may,perhaps, infer that the pointed end of the fire-sticksis driven initothe ground diagonally,the otherend of the stickbeing held in positionby the two pegs whichare also forcedinto the ground in such a way as to crossonie another above the stick. One of the cane-thongswould then be passed underthe stick and the ends pulled alternatelyas usual. One of the fire-sticksexhibits a longitudinalgroove on one side,evidently produced by fire-makiingby the " ploughing" method, which appears from this to be an alternativemethod, perhaps derived fromcontact with Papuo-Melanesianculture, the iniflueniceof which extends westwardsas far as Cape Possessionat least, and nay well have reached the strictlyPapuian tribes still furtherto the westward. We know of the " ploughing"method of fire-makingin otherparts of the Papuan Gulf,extending as far west as Kiwai Island at the mouthof the Fly River. We next meet with the flexiblefire-saw in a groupof localitiesto the west of the extended Owen Stanley Ranige, in a region where the western Papuo-

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 46 HENRY BALFOUR.-FrictionalFire-makWing

Melanesian stock is more or less in direct,contact with the pure Papuanl,and where Negritic affinitieshave already been traced in certain districtsand seem likely to be noticedover a widerarea when the various tribeshave been subjected to more critical investigation. The Rev. Father V. MI.Egridil has describedand figuredthe process among the KUNI tribe (Kovio group) of Dilava, in the mountains behind Mekeo, c. 8? 33' S., 146? 52' E. These people speak a Melanesian language,though this is probablynot a safeguide to the racial affinities of this short,dark, frizzly-haired people. The Kuni usually keep their firesburning, and guard them jealously, and when travelling,carry a smoulderingpiece of softwood or of bark-cloth;but when firehas to be obtained,dry leaves are laid upon the ground,and over these some old bark-cloth,to catch the spark caused by sawing a cane stripround a stickheld upon the groundby the feet,until the cane breaks (Fig. 21). The operationof

IKuNI,DILAVA,,N KAMBISA CHIRIMA VALLEY,N GUINEA FIG. 21. FIG. 22. GUIlEA fire-mnakingis called vinW-jaka,i.e., to pull the cane thong (vine); the fire-stickis called aukapa. These names are only used when these materialsare employedfor fire-making,au and valo being the ordinarynames forthe wood and the cane. To the east of Dilava and on the south-eastslopes of Mount AlbertEdward, in the valley of the Chirima(a tributaryof the Mambare River),lie the KAMBISA villages at an altitude of about 6,000 feet,c. 147' 30' E., 8? 35' S. In this district Mr. C. A. W. Monckton in 1906 met with a people of short stature, allied, apparently,to the Kovio group,and thereforeto the Kuni. Their system of fire- rnakingis, like that of the lattertribe, by means of the flexible thong-saw. A piece of drywood is split in two, the two parts being bound togetherwith bark stripstowards one end (cf. Fig. 17), and forcedapart witha bamboopeg towardsthe otherend. This piece of wood is held downwith the foot. A bamboosawing-thong is passed underneathand round both limbs of the split billet, the ends being alternatelypulled, the spark is caught in tinder lodged at the point of friction2 (Fig. 22). It is to be noted that betweenthe Kuni and the Kambisa tribeslie the

1 Anthrovos,ii, p. 108. 2 C. A. W. Monckton,Annual Report, Brit. New Guinea,1907, Plate accompanyingReport No. 108.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 47

Mafulu,recently studied by Mr. R. W. Williamson,whose researches have revealed a verystrong Negritic strain in this sectionof the Fuyuge-speakingpeople. Mr. Williamsonand othersthink it probablethat thisNegritic element is traceableover a wide area of the nmountainregion in this part of New Guinea, and the short statureof the Kuni and Kambisa tribes,to take one characteristiconly, would lend colourto the beliefthat these people also show pygmyaffinities.

OWE.N STANLEY RANGQ,N.G&UINEPi FIG. 22A.

A fire-stickfrom the Owen StanleyRange (exact localitynot specified),now in Mr. E. Bidwell's collection,is shownin Fig. 22A. The stick is, as usual, split at one end and wedgedopen with a peg. It has been sawn once with a strip of rattan about 6 mm. in diameter. I am indebted to Mr. Bidwell for particulars and sketches. At KAGI, on the western spurs of the Owen Stanley Range, under Mount Kaneve, c. 147? 40' E., 90 10' S., and at an elevationof 4,000 to 5,000 feet, fire is made with the same apparatus,according to Colonel Kenneth Mackay.' " They make fireby passinga lengthof cane undera partlysplit piece of soft stick (the cleft being held open by a pebble),and draw it quickly up and down where the wood restson some drymulberry leaves. In responseto the friction,smoke begins to rise in threeor fourminutes." Mr. StaniforthSmith (op. cit.) has drawnatten- tion to the similarityin physique observable between the Kagi people and the Sambregiof Mount Murray,who also practisethis methodof fireproduction. In the neighbourhoodof Kagi, i.e.,on the upper watersof the Naoro or Brown River and on the slopes of the main range,this processwas observedby Dr. H. 0. Forbes at Uburukara,situated on Mount Archer,which is due south of Mouint Owen Stanley,c. 147? 32' E., 90 0' S.2 He says, "Most of the men wore on their arms,or hangingin theirgirdles behind, coils of rattanrope, which are used by them for'making fires.' For this purpose the operator,first selecting a dryfraginent of wood,makes in it a split,in whichhe insertsa peg to keep it agape; into this split he places looselya morselof tinder plucked out ofhis girdleor skirt. He nextcuts fromhis drycoil of rattana shortlength, lays it on a dry leaf on the groundalid places over it the tinderplug in the cleftstick; thenplacing his knee or footon the end of the stick,he pulls the rattancord rapidly to and frounder it till the tinder ignites,when, by blowinggently through the cleft,he fans the sparkinto a flamie. The whole operationis the most effectiveand rapid of any native fire-producing contrivancesthat I know."' On the lower slopes of the main Owen Stanley Range, in the hinterlandof Port Moresbyand the Motu area, the KOIARI employthe same device,as observed

1 AcrossPapua, 1909, p. 149. 2 Proc. Geogn.Soc., xii, 1890, p. 562.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 48 HENRY BALFOUR.-FrictionalFire-making by the Rev. W. G. Lawes in the Tabure and Sogere villages on the slopes of Mount Warirata(Astrolabe Range), c. 90 25' S., 147' 20' to 147? 33' E., the natives of whichare typicalKoiarians.1 " They get a drystick of pithywood and splitit a little way. In the cleftthey put a piece of wood or a stone to keep it open, then puttilnga little rubbish as, tinderunder the split part of the stick,they stand onlthe otherend and pass a strip of rattancane or bamboounder the cleft,drawing npit rapidly and down,when it soon beginsto smoke,and a spark appearsbetween the forkof the stick,which with a little manipulationsets fireto the tinderand a flameis soon obtained. It seems to me easier and quicker than the commonway of gettingfire with two sticks." Dr. 0. Finsch also saw firemade in the same manneramongst the Koiari. He tells us2 that the stick,called ntewata,is short,peeled and lonigitudinallysplit at onle end, the cleftbeing wedged open with a stone. The native takes a handful of dry grass,rubs it up into a pellet and lays it underthe cleftstick, on whichhe places his footto hold it firmly. A long stripof split bamaboo,called ana, is drawnto anld fro throuighthe cleft, fire often being obtainied in - ~ ) 30 seconds. The atnc is carried by everynative; any

____ _ dry stick will serve for the " hearth." It will be noted that the thongis describedas passing throughthe cleft, but the figure(Fig. 23) given by L. Frobenius3of the apparatus which Finsch gave to the Berlin Museum showsthe thongpassing round bothlimbs of the cleft N. 1KOIAF8 G~tJ~Nstick, in the more usual manner,and, moreover,the FIG. 23. friction-groovesappear to indicatethat this was actually the case. It seems likely that Finsch's descriptionwas slightlyat fault,in spite of there being precedentin Borneo for the procedureas described by him,and in GermanNew Guinea also,if Miklucho-Maclay'sdescription, quoted below,is correct upon this point. In GermanNew Guinea the occurrenceof the flexiblefire-saw has been noted by several observers,though the recordedlocalities are all confinedto the eastern districtlying between Astrol.abe Bay and Huon Gulf,and are situatedeither on the coast or at no great distanceinland. The firstpersoni to refer to this processin GermanNew Guiineawas the Russian, N. v. Miklucho-Maclay,4who observedit in 1872 in some of the hill-villages of Enalam-inala and Tieingurn-manlanear the Malay Coast (Rhaj Coast), to the,south of AstrolabeBay. A piece of very dry wood,called ilol, was split at one end with a stoneaxe, and a stripof split cane or liana was passed through the cleft and sawn backwards and forwards with inereasinigrapidity in the usual manlner. The woodenbillet was held down on the

1 Letter fronmMr. Lawes to the Roy. Geograph. Soc., Proc. B.G.S., v, 1883, p. 357. 2 Ann.d. K.K. NaturI&ist.Hof-museums in W7ien,iii, 1888,p. 323. 3 V&lkerkundein Characterbildern,1902, i, p. 342. 4 "Ethn. Bem. uiberdie Papuas," YVatuurkund4jetijdsckrift voor lVederlandsch-Indie, xxxv, 1875, p. 83. Quoted in B. Hagen, " Unter den Papuas," 1899, p. 204.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 49 groundwith the knee or the foot. Tinderof coconut-fibrewas placed under the frictionpoint. The operationlasted half an hour. Here, again, the thong is describedas passino throughthe cleft,and olnewould have liked confirmationof this. The time stated to be required to obtain the spark is remarkablein view of the statementsby other observersas to the ease and rapiditywith whichfire can be procured by this method. One wondershow a thin stripof cane could withstand the frictionso long withoutbreaking, and also how the operatorcould continuethe motionfor half an hour withoutbecoming exhausted. Half a myiinuteis nearer the timeusually requiredto producethe desiredeffect. Further south, the method is employed by the POUM,who dwell in the mountainsinlaiid fromthe coast,to the north-westof Finsch-Hafen. The process has been describedby Dr. Rudolf Pbch,lto whomit was shownby Mazeng, one of the chiefsof the tribe,who selected a dry sticI' with the bark attached, about a. metrelong and about 5 cm. in diameter. This was split longitudiniallyat one end,,

Pour1 GCeRMAN N GUINEA

FIG. 24. the cleftbeinig wedged open with a small piece of wood (Fig. 24). This stick was boundround to preveiitthe cleftextending too far. It was then firmlybound to the house-postin a horizonitalpositionl. Next, Mazeng,tore a fragmenitfrolm his bark coat and forcedit into the cleft. Then he uintwistedone of the manyarmlets of coiled rattanstrip which lhe was wearing,so as to have a thongabout a metre in length. He and another ( / '/i " masi, standin on oppositeft sides of the fixed clefte stik, laid the cane thong over the wood where the tinder was placed, and Po u M pulled the ends alternately FIG. 25. until, at the end of not more thaii a miinute,the caine snapped, by which time the tiniderhad blecomeigniited (Fig. 25).2 P6ch is no doubt right in

1 Mit.d. Anthrop.(Ges. in WRien,xxxvii, 1907, p. 59; Globus,xcii, 1907, p. 196; Geographical Journal, xxx, 1907, p. 612. 2 This figure,copied from Leitfaden der VUlkerkunde,by Dr. K. Weule, is apparently based upon Poch's description. VOL. XLIV. K

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 50 HENRY BALFOUR.-FrictionalFire-making suggestingthat the wearingof this particular kind of cane armlet is connected with this methodof makiilgfire. It is verycommonly worn amongst the Poum, but is rare among the neighbouringKai tribeof the Sattelbergdistrict, who do not appear to make firein this manner. The cane thongis always readyto hand,as it is oily necessaryto untwista convenientlength from the armlet,and, as has been pointedout, the Tapiro pygmiesof Dutch New Guinea wear identical armlets and employthem in an identicalmanner for fire-making. The Poum do not exhibitany markedNearitic affinity, but it is noteworthythat the Kai, thoughspeaking a pure Papuan language,are a mlleso-to brachycephalicpeople of low stature, the men averaging5 feet,while a considerablepercentage measure only 4 feet4 inches.1 So that it appears that a pygmyelement remains in this region. The HUPE, who are near neighboursof the Pouni, livinigto the south-east of the latter,in a districtto the west of the Sattelberg,inland from Finsch-Hafen (c. 1470 30' E., 6? 20' S.), producetheir fire in a mannernearly identical with that practisedby the Poum. The followingdescription is derivedfrom Dr. R. Neuhauss.2 Among the Hupe, (as among the Poum, two men performthe operation. The cleft 8tiekis lashed to two otherlong sticksso as to forma tripod. Dry grass or bast is lodged in the cleftand two men saw a rattanthong across the cleftover the tinder, pullinigthe ends alternately(Fig. 26). Fire is quicklyand easily obtainedby tlhem. Neuhauss cinematographedthe process. In orderto have a ready supply of cane strip for fire-making,the iupe wear armlets of coiled cane like those of the Poum and Tapiro.

I NLAND FROMSAMOA HAFEN, C N G. HUPE, CXERMAt4 N &U NE-A FIG. 26. FIG. 27.

Lastly, the natives of the districtinland from Samoa Hafen, in Huon Gulf (c. 70 S.), as observedalso by Neuhauss (op. cit.),employ the cane fire-sawafter the less elaboratefashion prevalent in Britishand Dutch New Guinea. They do niot

1 GeographicalJournal, xxx, p. 610. 2 DeutschNeu Guinea,1911, i, p. 252,and iii, p. 24.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a FlexibleSawing-thong. 51 fixthe stickupon a framework,but the operator (one only) stands upon a stick whichis cleftand holds somedry grass as tinder,and saws with the cane thong passed underit (Fig. 27). The neighbouringLae-Womnba on the lower Markham River employthe " fire-plough." The referenceswhich I have so fargiven show a verywide, though undoubtedly conqnected,geographical range forthis specialized system of fireproduction, extendino, as it does,from Assam to Eastern New Guinea, the two extremesof dispersalbeing linked by sporadic occurrencesof this method throughoutthe interveningarea. There can be little doubt that the flexible thong-sawis closely related to the betterknown sawing method, i.e., with a rigid,blade-like " saw " of bambooor wood, whichprevails in the Asiatic Islands as far east as the Philippines and the Kei Islands, and to the westwardextends to the Malay Peninsula, Burma, Siam, the Nicobars and India (Madras Presidency). In Australia, too, this method of " sawing" fireis fairlycommon. In some instances,where wood is used instead of bamboo, the lower stick is split and the cleft is wedged open, just as is so very frequentlydone where the cane thong-saw is employed. Or, instead of splittingthe wood,a piece may be chosenwhich has a naturalcrack in it,answering the purposeof a cleft,or again a grooveor pit is cut in the wood and the "saw " is workedacross this. This use of a cleft or longitudinalgroove across which the sawinigis performed,affords an additional link between the two methodsof fire- sawing. It is difficultto determine which of the two methods is the more primitive,but perhaps the rigidsaw may claim to be regardedas the earlier. It is rathermore simple,and its prevalencein Australia.where the flexiblesaw does rnotexist, may emphasizethat clairnto priorityof invention. On1the other hand, the flexiblesaw is a markedfeature in the cultureof those tribes in New Guinea which showvNegritic affinities,or which are adjacent to Negritictribes, amongst whom the rigid fire-sawis not known. The one method has not been, strictly speaking,evolved from the other, since intermediatestages do not occur, and cannot very well do so; but it seems likely that both are variants of the scame -procedure,arrived at in slightly differentways and converging. A previous -knowledgeof the rigidfire-saw would readilyhave suggestedthe adoption of the flexiblesaw as a variant or alternativein districtswhere cane is commonand is used forother purposes in the formof narrowstrips. AFRICA.-Leavino,now this importantand extensive eastern area of dispersal of the flexiblefire-saw, let me turn to otherregions in whichthis particular manner of producingfire has been observed. As far as I am aware, there is only one otherpart of the worldwhere this processhas been noted amongsta people living under conditions of strictlyprimitive culture and this, curiously enough, is completely isolated from the rest of the geographical distribution. This is in West Africa,among the Ba-Kele (Ba-Kalai or A-Kele) of the Ogowe6 River district. The late Miss Mary Kingsleywrote,' " On one occasionI saw a Ba-Kele

' Travelsin WVestAfrica, 1897, p. 600. E 2

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 52 HENRY BALFOUR.-FrictionalFire-mnaki'g womanmake fireby means of a slip of raphia palm drawn veryrapidly, to and fro acrossa notchin anotherpiece of raphia wood." As far as I am aware,there is no othersimilar reference to this methodin any part of Africa,and one must regret that Miss Kingsley'sdescription of the processis so meagre,as it barelysuffices to convinceone that this is an undoubted instance of the employnmentof a flexible saw in fire-making.Anl illustration would have made it clear, but this, un- fortunately,is lacking. The expression" a slip of raphia palm drawn. . . to and froacross ....," seems definitelyto suggest a narrowand thereforepliaint band of the material,the ends ofwhich were alternately pulled, to cause the frictionby sawing transverselyacross the stick, and if this is the correct diagnosis, thenl the apparatusis strictlyanalogous to the flexiblefire-saw of the Indonesian area and of New Guinea. Moreover,the descriptioncannot apply either to the "hand- drilling" method or to the "ploughing" (" stick and groove") method of fire- making, which are the only other frictionalmethods described from"Savage" Africa,unless a statement,which has, I believe,been made, that a formof fire-saw is used in Katanga, is substantiated. Perhaps somelight may be thrownupon the Ba-Kalai fire-sawby other witnesses. In the meanitimeI merely referwith a certain amounitof reserve,due to lack of detail and corroborationl,to the only recordI have found of this characteristicallyeastern method being practised in Africa. EUROPE.-It is interestingoto note that amongseveral other methodswhereby "New-fire" or " Need-fire" is, or has been,produced for ceremonialpurposes, the flexiblefire-saw finds a place. I need not here enterinto the generalconsideration of the survival in Europe of primitive frictionalfire-making appliances, as a featurein certain religiousor superstitiousobservances. The beliefthat fireafter prolongeduse becoinesstale, as it were,and loses its mysticpower and virtues,is, or at any rate has been,very widely spread in Europe, even amid environmeintsof highly developed culture. The persistenceof this belief into quite recenttimes is veryremarkable.' The actual procedurevaries in the differentlocalities, and the apparatus employedto elicit the niew-fireoccurs in several differentforms, but I will confinemyself now to the instances of the employmentin Europe of the flexiblefire-saw for this purpose. SWEDEN.-The followingdescriptions and figuresof the variousprocedures as practised in Sweden, when need-fire(nodeld or gnideld)was required,are culled froma veryinteresting paper by Nils Keyland.2 First method.-A length of rope, willow-stripor strap is passed completely rounda log of wood, and the ends are pulled alternatelyby two men until a spark is generatedand can be caught upon a piece of tinder(Fig. 28). This imiannerof making fire is employed in Eksharad, Vermland, both for amusement and

1 Vide Grimm,Teutonic Xythology; Tylor, Early Hist. of Yankind; Kuhn, Die Herab- kunftdes Feuers,etc. 2 "Primitiva eldgdrningsmetoder i Sverige,"Fatabutren, 1912, pp. 15-17, anld 1913, pp. 92-94.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a PlexcibleSawing-thong. 53 ceremonially,in cases of special need. AugustBjornander, of Manstads socken, Kindshuirad,Vesterg6tland, who was born in 1839, describesa similarprocess, in which a bast rope is wound twiceround a drybranch of birchor aspen,on which the barkremains, or else rounda drytree-stump. The cord was pulled by two men.

VERMLPNO. 5WEDEN. VERMLSPND, 5WEoEN. FIG. 28. FIG. 29.

Secondmethod.-This varies from the firstonly in the thongbeing passed half r-o2&ndthe logfor stump. One man can thus performthe operation,pulling both ends of the thonghimself, though in the illustration(Fig. 29), two are shown,each of whompulls one end; Eksharad,Vermland. Accordingto informationsupplied by Nils Olsson Lack of Lindhojden,Elga socken,Verinland, need-fire was produced by pullinga cord passed half-rounda drypine, or fir-stick,which was held firmly between the knees (Fig. 30). This method was practised by the forestpeasants when necessitydrove them to it.

VERAMLAND.SWEUEN. VERM LAND. SWEDEN FIG. 30. FIG. 31. Third rnethod.-JanOlsson of Vaxvik, K6la socken,Vermland, described a miannerof producingneed-fire by sawinga rope backwards and forwardsover the

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 54 HENRY BALFOUR.-FrictionalFire-making top rail of a wooden fence (Fig. 31). The informationwas based upon hearsay, and it was not statedwhether the processwas conductedfor any special purpose. Fourthmethod.-This is a peculiar variant,as the thongis fixedand the billet of wood is moved along it. One end of the thong,which is passed rounda small stick, is attached to a stump,or other fixedobject, and the other end is held by a man who strainsit tightly,so that it gripsthe centreof the stick,which is thus held transversely. A second person, holding the ends of the stick in his hands, forces it backwards and forwards(spokeshave fashion) along the thong, the encircling portion of which causes enough frictionto generate the desired spark (Fig. 32), Eksharad,Vermland. This inversionof the more usual practiceis comparable with the invertedsawing processin Burma,Borneo and elsewherein the East, in whichthe bamboo blade (or saw) is fixed,and the "hearth" is worked

A;~~~~~~~~~~~~m

VE RMLA4D,SWEOEN F1G. 32. FIG. 33. like a spokeshave along its upturned edge, a methodsometimes practised as an alternativeto the more usual one of sawing the blade across the fixed" hearth." Both ways are equally easy when bamboois employed,and I have veryfrequently perforrnedthe operationsuccessfully. Fifth method.-Anothervariant of the flexible-sawmechanism was used in Smaland, South Sweden, in the eighteenthcentury (1763) for avertingthe evil glance (fascinationenb)when fishing-gearwas thoughtto have been bewitched. A hole was bored in an oaken door-post,and throughit a strap or cord was drawn backwards and forwardstill fire was obtained and could be caughtin brimstone. The fishing-gearwas then smokedover the fireso created. Accordingto Smiback-Johannesof Vaxvik,Vermland, a mianof about80 years of age, need-firewas made by a similar procedure in order to cure pigs when smitten by disease. A leathern thong was passed througha hole in a door,and a couple of men,standing on either side of the door, pulled the ends alternately until a spark was generatedby the frictioln(Fig. 33). From this a fireof shavings, etc.,was kindledat the door of the pig-stye,and the pigs were driventhrough it.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a FlexibleSawing-thong. 55

Sixthmethod.-The descriptionof a verycurious and seeminglyunique thong- sawing method of fire-productionis contributedby J. P. Hedberg,who says that his father,who was born in 1828, served in his youth as farm labourerwith a peasant in Abyn,Savar socken,Vesterbotten, North Sweden. One day he and his employer were in the forestand wanted a fire, but had no means of making it. So theytook a fairlylonlg strip of withy or birch which they twistedup into the formof a closed ring. This ring or loop theyfitted round a dry old fir-stump,v and theythen pushed the end of a stick between the stump and the ring,twisting the loop once ,,4 jf4 > round the stickto make it grip the stump firmly, --. whereupon,using the stick as a lever,they walked ANGERMANLAND,SWEDEN round the tree-truink(Fig. 34). The frictionof FIG. 34. the cord,thus sawn against the stump generatedsmoke, and later a spark,from whichfire was obtained. J. P. Hedberg,although he was not presenton the above occasion,at a later date saw fireproduced for fun by the processwhich he describes. This was apparentlyin Angermanland,North Sweden. GERMANY.-Grin-mlquotes the followingpassage fromLindenbrog's Glossary of thleCapitularies:-" Rusticani hominesin multisGermaniae locis, et festo quiden?, S. JohannisBaptistae die, palumn sepi extrahunt, extracto funem circumligant, illumque hu6eitlluc ducunt, donec ignernlconcipiat, quern? stipula lignisquearidioribus aggestis curatefovent, ac. cinerescollectos supra olera spargunt,hoc medioerucas abigi pQsse inane superstitionecredentes. Eum ergo ignem " nodfeur<" et " noclfyr,"quasi necessariumignem^, vocant." The descriptiongiven in this account appears clearly to indicate a processof obtainingnew-fire (nothfeuer of the Germans)by pulling a cord to and fro round a stake and causing a spark by sawing-friction.The expression" illumque huc illutcducunt " seems moreapplicable to drawingthe corcl (funis) than to turningthe stake (palus), though the possibilityoccurs to one that, this recordis merelya careless descriptionof the farcommoner method of turning a spindle,jammed between two uprights,with the help of an encirclingcord, after the fashionof a " thona-drill." The latter processwas widelypractised in Europe, from Scotland and Scandiiiavia to the Balkan States. We may, however,be contentto take the passage literallyand accept it as affordingan instanceof the use of the flexible-sawingmethod in Germany; the more so, since there is precedentfor its occurrencein variousparts of NorthernEuLrope. The particular use of the ashes of the fire kindled in this manner-that is, for drivingaway caterpillarsfrom the kitchengarden-is interesting. In the Mark, Germany,the customprevailed of makinga nothfeuteron certain occasions,especially when there was disease amongthe swine. "Before sunrise two stales of dry wood are dug into the ground amid solemn silence,and hempen

1 TeutonicMytholo,y, 1883, ii, p. 604.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 056 HENRY BALFOUR.-FrictionalFire-mak-ing ropes that go round them are pulled back and forwardstill the wood catches fire; the fireis fedwith leaves and twigs,aiid the sick animalsare driventhrough."' Here, again,a suspicionmay arise that the apparatus may have really been of the nature of a " thong-drill"rather than a " cord-saw." At least, the referenceto two stakes fixed in the ground is suggestiveof the two uprightsbetween which a -spindlewould be janimed, so that it could be rotated with a cord and create frictionagainst the uprights. It is regrettablethat the descriptionshave not furnishedmore detail, makingthe precise methodperfectly clear. The practiceof kindling nothfeuerby frictioniconltinued in the Mark as late as the middleof the nineteenthcentury. RUSSIA.-An interesting account of procuring need-firefor prophylactic purposes,is given in a chapteron " leligio RussorumAliae superstitiones,"written in 1581.2 1 am indebted to Sir A. Evans forthe loan of this book. The passage runs as follows:-Luemn castr-ensen?,quam nostrimilites huc advexerant,tali modo eospropqtlsare vidi. Funem accipiunt,eumque per foramiien scamni in domovetustioris tantispertrahunt et retrahunt.donec ignis scintillamnex arido assereeliceant: flanmma conspecta,cereumn in delubroundis lustralibuspegfusum accendunt. Omnesdeinde qui ex hac favilla prunas vel torremtdo?ni conservant,tutos fore credunta pestife'ra contagione,et fit nonnunquatm,-ut eventusfidei respondeat." The process here describedis identical with that already referredto, as occurringin Smaland, in Southern Sweden. A cord was drawn to and fro througha hole in a dry old bench, and from the fire obtained by the frictiona wax taper,which had been sprinkledwith holy water,was lighted. Embers kindled fromthis and preserved in the houses, were believed to protect the people from the conitagionof camp- fever; and, as the autlhordryly remarks, "It sometimesdoes happen that the result is in keeping with theirbelief." From the contextit may be inferredthat the above descriptionapplied to Ruthenia(Little Russia). I have now quoted the variousreferences to whichI have had access, dealing with the processof fire-makingwith a flexiblesawing-thong, and it will be seen thatthere are threedistinct areas of geographicaldistribiution. (1) An Easterii area extendingfrom Assam to New Guinea; (2)West Africa;(3)Europe. These three areas must be consideredseparately, as, in the absence of evidence which might indicate a culture-connectionlinking them together, we must,provisionally at any rate, regard the use of the sawing-thongin fire-makingas havingvery possibly ,arisenindependently in the threeregions. The firstof these,the EASTERN AREA, in spite of the manygaps which break the continuityifn distribution of this appliance at the presentday, none the less appears very definitelyto representa connectedculture-area. Within the limitsof

I AdabertKuhn, fdrrkischeSagen unidAidrohen, 1843, p. 369,quoted in Grimm'sTeutonic Mythology,II, p. 605. 2 JohaninesLasitzki (Lacicius Polonus), "De Russorum,Moscovitaruni et Tartarorum religione,sacrificiis, nuptiarum, funerum, ritu. E diversis scriptoribus." Spirae Civitate VetergumNemetum, excudebat Barnardus D'albinus, anno 1582. Page 246.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a Flexible Sawing-thong. 57 the easternrange of the thong-saw,there are traceableseveral widely differentiated racial elements, though the purity of type is usually greatly obscured by miscegenation,a factwhich renders very complicatedthe ethnologicalproblem in this region. An ethnologicalstudy of the flexible fire-sawseems to bring into prominencetwo of these racial elemeentsmore especially,-the Indonesian (pre- Malayan) and the Negriticstocks-and it appears probablethat to one of thesethe originand inventionof this varietyof the fire-sawis due. Although there are a fewvariations in mattersof detail in the mode of employinentof the apparatus, these are not sufficientlydistinct or localized to warranta theoryof independent inventionin two or morecentres within the Eastern area. As has been seen from the descriptionsgiven above, these variationsare nlotas a rule restrictedto any special districts,but are liable to recur even in widely separated regions; while such instancesas occur of local specialized types-such as the anthropomorphic design of the Kayan fire-stick-areclearly due to an elaborationof thenormal and widely-spreadform of the apparatus. At one extremityof the rangeof dispersalin the East are the Tibeto-Burman Nagas and the nearly-relatedTipperahs, and they,as well as the Mois of tlhe Annamufrontier, must be linked cultuirallywith the Indonesianpeoples. Thereare verystrong affinities between the cultureof the Nagas alnd that of some of the natives of the Asiatic Islands, notably the natives of Borneo; while striking similaritiesm-ay be tracedfrom Assam as far east as the Philippinesand Melanesia, and evenifurther. The coincidencesare too numerousto be fortuitous. In the Malay Peninsula,which, presumably, was on the line of route of the early migrationfrom the Asiatic mainland into the Archipelago,we findthe Sakai, Semang and Malays all employingthe thong-saw. The short-statured,curly-haired Sakai (pre-Dravidian or Dravido-Australoid)appear to have racial affinitieswith somieprimitive peoples in the Asiatic Islands (e.g.,Sumatra and Celebes),and even, accordingto Moszkowski,in New Guinea (Geelvink Bay). The Negritic Semang are akin to the Negriticpygmies, of whomtraces may still be foundin severalof the islands of the Archipelago,and who persist,as a remnantof an early imiimigration, in New Guinea, and still forman importantelement in the local population. It appears likely that the Malays of the Peninsula derivedthe thong-sawingprocess eitherfrom the Sakai or the Semang,since, as faras I am aware, thereis lnorecord of this methodof fire-productionin Sumatra,the originalhome of the Malays,nor among true Malays elsewhere. In Borneo,the Kayans, accordingto Hose, probablyrepresent an offshootfronm the Indonesianmigration from the mainland,which spread froomthe basin of the Irrawadi,and theyhave physicalas well as cultural affinitieswith the Nagas. The Kenyahs are believed to have been a still earlier immigrantpeople exhibiting a blendingof Caucasian and Mongoloidelements, who enteredBorneo whileit was still conlnectedwith the mainland. Hose is also of opiniolnthat the Kenyahs lhave derived the principal elements in their culture fronmthe Kayans, whose characteristicsthey have assim-ilatedvery completely. Probably the use of the

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 58 HENRY BALFOuR.-FrictionalFire-making thong-sawby these two tribes almost exclusively among the inhabitantsof the island is referableto this transmissionof culturalelements from the one tribe to the other,and in this case it is likely that this apparatus reached Central Borneo and Sarawak fromthe S. or S.E. partsof theisland, and accompaniedthe migration of the Kayans along the largerrivers into the uplands of the centralregion.' The formerexistence of a Negriticpeople in Borneo,which althoughnot colnclusively proved,is regardedas highlyprobable, gives rise to the alternativepossibility of the Kayans and Kenyahs having derived the process from an earlier pygmy population,akin to the Semang,and of whomvery fewtraces remainto-day. In the Philipyinies,the flexiblefire-saw appears to be associated,mainly at any rate,with tribesof Negriticor partiallyNegritic stock, though the pre-Malayan, Indonesianelement is widelyrepreseiited in the group,strongly tinged with Mon- goloid characteristics;and it may be regardedas an open question,whether the thong-sawwas introducedinto the Philippinesby Indolnesiansor by Negritos. The passagefrom Borneo to the main Philippinegroup via Palawan and Minidorowould have been anieasy one, the distancesby sea beingshort. When we come to New Guinea, the association of this apparatus with the pygmyNegi,itic race becomesmore marked. Althoughfrom the descriptions given in this paper it is clear that in certaindistricts tribes of pure Papuans, or, in some cases, of Papuo-Melanesianstock, practise fire-making by this process,it must be admittedthat the thong-sawis moreparticularly characteristic of the uplandpygmy peoples,and of thosetribes in whose physiquesome Negriticfeatures are traceable. It is noteworthythat, with comyiparativelyfew exceptions,the Papuan and Papuo- Melanesian empLoyersof this systeinof fire-makino,inhabit districts where they have been liable to comeinto contact with peoples whoseNegritic affinities are more or less well defined; and in view of the frequenitoccurrence of thethong-saw among the upland peoples,usually of low statureand presentingother Negritic tendencies, and also of the factthat the neighbouringtaller Papuans and Papuo-Melanesians moreusually adopt some othermethod (e.g., the " fire-plough")-it appears probable that the flexible.fire-sawwas introducedinto New Guinea by the early immigrant Negritos,and was transmittedby themto some of the later dominantpeoples, as a result of culture-contact. Apart fromthe generalclose resemblanceobservable in the apparatusand in its mode of employmentin the two regions,the use of a cane stripin verylong lengths,ordinarily kept coiled up forwearing as armlets,affords a strikinglink betweenthe thong-sawingprocess in the Philippinesand its counter- part in New Guinea,in both of whichregions this methodof fire-makin,is closely identifiedwith Negrito culture. The finalsolution of this problemmust await furtherinvestigation. At one (the north-west)end of the range of its dispersalin the east, the probabilityof an Indonesianorigin of this contrivanceis distinctlysuggested; while, on the other

I Hose and McDougall, The Pagan Tribes of Borneo,chapter on the ethnologyof Borneo.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a FlexibleSawing-thontg. 59 hand, at the other (the east and north-east)end, the Negrito origin is equally definitelyindicated. A wider knowledgeof Negritoculture and distributionmay enlightenus. Both the split and the unsplitforms of the fire-stickor " hearth" occurat the extremeends of the distribution,and the details of method in the employmentof the apparatus,while varyilnglocally to some extent,do not present such well-definedvariations in the several districts,as to offerany definiteclue as to the originalhome of the system. It seems probable that the flexiblefire-saw of the East was invented as a variant from the better-knownrigid, blade-like fire-saw. It occurs to a great extentover the same area as is occupiedby the latter,the two processesbeing sometimesemployed alternatively by the same people. Bamboo,which is the prevailing materialemployed for both parts of the rigidfile-saw, is used, instead of wood,for the " hearth" of the thong-sawby certaintribes (e.g., the Mois and Tipperahs)and occasionallyby the Sakai, and this affordsa link between the two methods. A furtherlink betweenthe sawing-bladeand the sawing-thongmethods is suggested by the occasional splittingof the wooden"hearth" sometimesemployed with the former,as, forinstanice, in Australia. The second distinctarea of distributionof the thong-saw-i.e.,WEST AFRICA- is a peculiarlyisolated one, since, as already stated,the only people in Africawho have been creditedwith the use of this method of fire-makingare the Ba-Kalai of the Ogowe River district. The possibilityof an independentorigin of the system in this region is forciblysuggested. At the samnetime, when we recall tlle numerousstriking resemblances observable in the arts anid appliances of native tribesin Africa on the one hand,and thoseof the Southern Asiatic area and the Malayan Archipelago on the other, it inust be admitted that it is not quite impossiblethat the Ba-Kalai may owe theiracquaintanceship with the thong-sawto somleremote contact with the sources of Asiatic culture, and that there may perhaps,be a true phylogeneticrelatioiiship between these now widely separated formsof the apparatus. In connectionwith the Asiatic and Indonesian elements traceablein Africanculture, I can only here referto Dr. L. Frobenius'book' and to a paperby Dr. B. Ankermann.2 The generalsubject is too wide forme to eniterinto it here. The third area to be briefly considered-the EUROPEAN area-is again apparelntlyvery distinct from and quite unconnectedwith the other two, as faras the appliance under discussionis concerned. Thereseems little likelihoodof tracingin Europe such affinitieswith Negritic,Indonesian, pre-Malayanor Bantu culture,as would warrantthe suggestionof a probablecommon origin for the thong-sawof Europe and that of other parts. As far as the records show, the variety of frictionalfire-production in Europe is and has been associatedalmost, if not quite exclusively,with the practice of procuringceremonial-or need-fire,but for this

1 Ursp}ung dcer IKultur, 1898, vol. i. 2 " Kulturkreiseund Kultursehiebtenin Africa,"in Zeit.f. Ethnologie,1905, p. 54.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 60 HENRY BALFOUR.-FrictionalFire-makinLg purpose-European peopleshave farmore frequentlyresorted to the thong-diilling method,whereby a spindleis caused to rotatein a socketcuit in a verticalstake, by means of a cord or thongpassed onlceor twiceround it, the ends of the thongbeing, alternatelypulled so as to cause a reciprocalrotation of the spindle,the frictionof whichagainst the stake generatesthe heat requiredto createa spark. This thong- drillingprocess is widelydispersed in Europe (BritishIslands, Scandinavia, Switzer- land, Germany,the Balkans, etc.), and it has occurredto me as verypossible, that it must frequentlyhave happenedthat, owing to the spindle becomingjammed, it may have refused to rotatewhen the thong was pulled. In such an event, the thong would have slipped on the fixedspindle as its enldswere violentlytugged alternately,and the frictioncreated by thisinvoluntary sawing of the cord upon the wood,would have caused smoke to arise in a very few seconds, therebyindicating that firecould easily be obtainedby this method,as an alternativTeto employing the rotatingspindle. The misuse of the one methodmay so readilyhave suggested the other,that a theoryof the likelihoodof the thong-sawhaving arisen indepen- dentlyin Europe may be regardedas a perfectlyplausible one. In conclusion,let me say that whereasI have endeavouredto bring,together as manyreferences to the use of the thong-sawingmethod as possible,I almneverthe- less consciousof the factthat my list is probablyfar from complete, and that inuch moreinformation and materialis probablyavailable. Many unrecordedexamples of the apparatusmust exist in museums,and thereare, no doubt,miany references in literaturewhich I have not yet seenl. I shall feel verygrateful for any information whichwill help in tracingstill furtherthe geographicaldistribution and varieties of this interestingfrictional fire-making process, and I trustthat at a later stage manyof the gaps in the continuityof dispersalnmay be filledup, and sufficientlinks in the seriesbe forthcomiingto admit of a mole completeand detailed treatmentof the subject.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions wvitha Flexible Sawing-thong. 61

M A S E.E.AsIA eTHE ASIATIC. I SLAND5. r t27 Ib? I2O

| M>XX~~~~~~~~~~~~~~~~~IUA

MAP I.

J E . 50 AXR G.N A

MAP II.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 62 HENRY BALFOIJR.-Fqictional Fir e-making

MAP 1III.

MAP IV.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions with a FlexbTleSawing-thaong. 63

LIST OF THE FIGURES ILLUSTRATING THIS PAPER. 1. Naga of the Naga Hills, Assam,making fire with the thong-sawof cane and a dry log. AfterS. E. Peal, Journ.Anthrop. Inst., xxii, 1893,Plate XVIII. 2. Naga-thong-sawingapparatus of wood and cane. From a roughsketch of a specimenin the Berlin Museum. 2A. Naga thong-sawingapparatus from Naga Hills near Mokokchung,given to the Pitt Rivers Museum,Oxford, by Mr. J. HI. Hutton. a = the fire-stickof lime-woodshowing one used fire-grooveand notchescut as a startfor the next; b = a bamboo strip fromwhich the narrowsawing-thongs are split,about 54 cm. long; c = one of the unusedsawing-thongs trimmedup foruse; d = one of the thongswhich has been snappedin the middleby use; e the tinder made from the shavings produced in trimmingand smoothingthe thongs. 3. Senmangthong-sawing apparatus from Grit, Upper Perak. The stickof softwood is 59-6cmi. long. Collectedby Messrs.Annandale and Robinson. Now in the Pitt RiversMuseum. 3A. Thongfire-saw of the Semangof Grit,Upper Perak. BritishMuseum. Copiedfrom the " Handbook to the Ethnographical Collections," 1910, Fig. 7c. 4. Sakai of Perak makingfire with the thong-saw. From Ca,ptainG. B. Cerruti," Nel Paese dei Veleni,"1906, p. 161. 5. Thong-sawingapparatus of the Malays of Perak, collected by Mr. Leonard Wray. The stickis of verysoft wood and is 76 cm. long. Pitt RiversMuseum. 6. Malay of Perak sawing firewith cane thong. From a photographsent to me, by Mr. LeonardWrav. 7. Kenyah fire-sawingapparatus, Pata River,a tributaryof the Baram River,Sarawak. With bundleof bamboostrips for sawing. Used at the ceremonyof naminga child. From Dr. W. H. Furness," Home-lifeof the BorneoHead-hunters," 1902, p. 38. 8. Kayan of Sarawak makingfire with the thong-saw. From a photographgiven me by the late Mr. RobertShelford. 9. Kayan ceremonial fire-makingoutfit, Baram district,Sarawak. The " hearth " is 27 cm. long and is carved to represent Laki Pesong. Used at the child-naming ceremony. Cane thongs and the small mat uiponwhich the firewas made are tied up with the " hearth." Specimen given by Mr. C. Hose to the Pitt Rivers Museum. 10. " Hearth" of ceremonialfire-making apparatus, Baram district,Sarawak, of soft,white wood, 58-2 cm. long, carved at both ends; with cane sawing-thong. Specimen sent to me by Mr. C. Hose. 11. Kayan ceremonial fire-sawingoutfit, Barani, Sarawak. The " hearth " is split but is not carved; 29-8 cm. long. Cane strips and tinder of palm-scurf. Specimen given by Mr. C. Hose to the Pitt Rivers Museum. 12. Kayan ceremonial fire-sawing apparatus with carved "hearth," cane strips and small mat, Barain, Sarawak. From C. Hose and W. McDougall, "Pagan Tribes of Borneo," 1912, II, Plate 168, Fig. 2. 13. Fire-saw of the Kayans of Central (Dutch) Borneo. From a photograph by Dr. A. W. Nieuwenhuis, " In Central Borneo," 1900, I, Plate 19. 14. Fire-sawingoutfit of the Negritos of Casiguran,North-East Luzon, Philippine Islands. a = the split " hearth "; b = the long, coiled cane thong; c = bark-cloth used as tinder. Copied from Dr. A. B. Meyer, " Die Philippinen," Dresden, 1893, Plate II, Fig. 7. 15. Pygmy of Palawan, Philippine Islands, making fire with rattan sawing-thong, bark cloth being placed under the fire-stickto serve as tinder. After a photograph by F. Cooper Cole, "Customs of the World," 1913, p. 657. 16. Fire-saw of the Tapiro pygmies of the Mimika River, Dutch New Guinea. a = the split fire-stick,25-3 cm. long, scored with many grooves from use; b = coiled up rattan-strip used as a sawing-thong. Specimen given to me by Mr. A. F. R. Wollaston.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions 64 FrictionalFire-making with a FlexibleSawing-thonzg.

17. Another example of the same, but the stick is split right through, the two parts being secured together with a cane strip. Given me by Mr. A. F. R. Wollaston. 18. Tapiro pygmy using the thong fire-saw. From a photograph in "Pygmies and Papuans," by A. F. R. Wollaston, 1912, p. 200. 19. Fire-stick probably from the head-waters of the Fly River, New Guinea, grooved by the frictionof a rattan strip. Copied from L. M. D'Albertis, " New Guinea," 1880, II, p. 378, Fig. 2. 20. Fire-making outfitof the Maipua tribe, Parari Delta, British New Guinea; tied up in a, bundle as carried about. Length of fire-stick,85 cm. The cane thong is in long strips. Specimen in the Pitt Rivers Museum. 21. Man of the Kuni tribe, Dilava, Mekeo district, British New Guinea, making fire with cane sawing-thong. From a photograph by P. Fillodeau, Anthropos,II, p. 108. 22. Fire-making apparatus, Kambisa tribe, Chirima Valley, Mount Albert Edward, British New Guiinea. From sketch by C. A. W. Monckton, Anu?ual Report,Brit. New Guinea, No. 108, 1907. 22A. Fire-stick from the Owen Stanley Range, British New Guinea, grooved by the friction of a cane thonig. Specimen in Mr. E. Bidwell's collection. Length about 29 cm. (the length was originally greater, but was reduced for convenience in packing), circumference7-5 cm. The sketch was not made directly from the specimienand mav not be perfectlyaccurate in detail. 23. Fire-making apparatus of the Koiari, inland from Port Moresby, British Museum, collected by Dr. 0. Finsch. Specimen in the Berlin Museum. Copied from Leo Frobenius, " V61kerkunde in Characterbildern," 1902, I, p. 342. 24. Fire-sawing apparatus with very long rattan thong, coiled up for wearing on the arm. Poum tribe, north-west of Finschhafen, German New Guinea. From a photograph published by Dr. Rudolf Pich, Alit. d. Anthrop. Gesellschaft in Wien,XXXVII, 1907, p. 59, Figs. 2 and 3. 25. Poum tribesmen making fire with the thong-saw. From Dr. K. Weule's " Leitfaden der V6lkerkunde," 1912, Plate CVIII, Fig. 9. The illustration is based upon the description given by P6ch. 26. Fire-making with the thong-saw. Hupe tribe, west of the Sattelberg, German New Guinea. From a photograph published by Dr. Neuhauss, " Deutsch Neu-Guinea," 1911, Fig. 170. 27. Native of the district inland from Samoa Hafen, Huon Gulf, German New Guinea, maaking fire with the thong-saw. From a photograph published by Dr. Neuhauss, op. cit. 28. Procuring " need-fire" with a thong-saw, Ekshdrad, Vermland, South-West Sweden. 29. Ditto, Eksharad. 30. Ditto, Vermland. 31. Ditto, Vermland. 32. Ditto, Ekshairad, Vermland. 33. Ditto, Vermland. 34. Ditto, Angermanland and Vesterbotten, North Sweden. The last seven illustrations have been copied from Nils Keyland's papers in Fataburen, 1912, Figs. 14, 15 and 16; and 1913, Figs. 20, 21, 22 and 23.

Map I.-Distribution of the thong fire-sawin South-East Asia and the Asiatic Islands. II. ,, ,, ,, ,, ,, New Guinea. III. ,, ,, ,, ,, ,, West Africa. IV. ,, ,, ,, ,, ,, Europe.

This content downloaded from 204.17.31.62 on Wed, 7 Jan 2015 14:55:50 PM All use subject to JSTOR Terms and Conditions