building a low impact roundhouse

8/22/2019 Building a Low Impact Roundhouse

1/121


2/121

Building aLOW IMPACT ROUNDHOUSE

Tony Wrench


3/121

Publishffi intheUKby:PermanentPubliu.tionsHyden House limited, The Susuinability Centre, East Meon, HampshireGU32 1HRTel: (01730)8233"Overseas:(int. code+-441730)Email:infopermacultuie.co.ukWeb: www.permaculture.co.ukUS distributors:Chelsea Green Publishing CompanyPO Box 428. White River Junction, VT 05001Tel:8o2 2956300Fax:8o22956444Web: www.chelseagreen .comTherightsofTonyWrenchtobeidentifiedasauthorofthisworkhasbeenassertedbyhiminaccordancewiththeCopyright,Des igns andPatentsAct988Text c o p y r i g h t ~ 2000 Tony WrenchFirstpublishffi in 2001Illustrations copyright 2000 John Adams I Martin Burton f Olwyn Pritchard ISpikeWatson J TonyWrenchDesignedandtypesetbyJohnAdams

PrintffibyBiddlesltd.,GuHdford,SurreyPrintedon]s%post consumerrecycledtotallychlorinefreepaperBritishtibraryCataloguing-in-Publicon ionData.Acatalogu-erecO


4/121

The Author

Tony Wrench has spent many years des igning and implementing,build ing and r e n e w ; ~ b l e energy projects. He lives with his mate janeFaith in the community at Brithdir Mawr, in Pembrokeshire,WestWalesTheir princip les are sustainability, simpl icity and spir it. Tony livesluxuriously, well below the poverty line, working on things permaculturaland wooden. Hemakes his living fromwood turning, singingand playingmusical instruments (some homemade) with the local circle dance andCeilidh band Rasalila.


5/121

lt is as naturalfor us to buildan appropriate shelter as it is

for badgers.

8 o d g u u n d e r a n o l d f i 1 d b o n t r u a t B r i t h d M a w


6/121

ContentsIntroduction . . .page 1 Windows & Doors ....... 63Backg round Basic Window Des ign .. 63Design ..................................4 Rubber Si ll Sheets ...........66Materials .. ................ ..10 Lintels .. .66

Embodied Energy Back Door .. . ......67The Skeleton .................. 5 South Windows .............68

Reciprocal Frame Roof Front Door ..... ..... ........70The Full Skeleton . ..........27 Stove Fitting .. . . ....n

lnner Uprights ..............16 Water --74Crosspieces ......... ........Jo Grey wat er --77Bracing . ....... .............. 1 Plumbing .. -79

The Roo f 33 Solar Panel ......................soSeconaary Rafters . ..... 35 Outside Touches .............83Willow Laterals .. 36 Earth Shel tering ..........83Mater ials ................ 37 Rubber Shingles ............ 85Roof Raising ...............38 Planting ....... ...............86Eaves/Roof Eage .............41 Electr ics ..........................89Turfin' .. .......... . . . . . ...... 44 wood ......................... 93The Sky l ight .... .. .........48 Compost Toilet 97

Walls 5, Construction ....................98Vents ... .... 59 Use .. . ... . .. ..........1ooCat Flap ........................6o Inside The Roundhouse .102Straw Bates .......... .... ...61 Endpiece ..........................107Platforms Et Partitions ....62 Appendix .. . ............ .. A1


7/121


8/121

IntroductionAre you superstitious? I may be taking a chance writing this. By the limeyou read i t our Roundhouse might have been demolished. ltmight havebeen listed as a historic building . lt might be frozen in asp ic as aSustainable lifestyle Heritage Exhibit with holograms doing repetitiouss ustainable activi ties in endless charade. lt might have collapsed! lt mighthave burnt to a cinder. Am I tempting fate by talking, nay, boasting,about it toyou?Time will tell. PermacultureMagazinepublished an articleabout this place in issue No 20, with drawings by Olwyn, a fr iend wholives lo u lly. Unwisely, I put at the e nd of t he article: 'for further detailswrite to Tony Wrench . ' I expected one or two letters asking " D id youtake the bark off the logs or not?" or "Did you use bolts on the s leepingplatform frame?" o r something. Instead I keep gening leners just saying'please send me more details'. OK, I give in. This book is the moredeta ils. Not everything. No fashion tips, recipes or laundry hints fo rhovel bui lders. I' ll try and put enough details in without boring you silly.Caution: Iam not even a cowboy builder. I'm not an architect or qualifiedto build anything. This place has more in common with a shack in ashanty town in Buenos Aires than it does with a new Wimpey house inEngland. Don't expect any expertise from this book. If you bui ldsometh ing like this roundhouse and it collapses, don ' t blame me. Thisis about doing things that are not regulated and predictable, I'm afraid.Naturally, I'd be delighted if you want to bui ld a roundhouse and usesome of the ideas he re, but if you' re wo rried about safety, consult anarchitect first. If you are a natura l worrier, better treat this as a bit ofsemi fiction, from another place and another t ime.


9/121

Background) ane and I have lived in West Wales since Dec 1989,trying out the ideasof permaculture on a smallholding of 1.5 acres, growing our own veg.,keeping chickens and ducks, living simply. We experimented wi th waysof reducing our bills by giving up a flushing loo in favour o fa composttoilet, and by building solar water heating panels and a rainwater showerheated by the sun. We converted the house light ing to 12 volts, installedphotovoltaic cells and a small wind generator, and I built and playedaround with various designs of den and studio buildings made fromrecycled materials (especially windscreens and double glazing units)and slab wood offcuts from the local sawmill. We got rid ofour car andbought an electric milk float. A normal country life, in fact. Each weekwe would hold a practice with Rasalila. the acoustic circle dance bandwe are still pa rt of, in the Cone, a wooden t ipi studio that doubled as ourbedroom. Emma, who plays fiddle, bagpipes, whistles and clarinet, usedto fill us in on progress at Brithdir Mawr. A tumbledown farmhouse andoutbuildings with acres of fields and forest about 18 miles from us, thatshe and her husband, julian, had bought wit h the intention offorming asustainable community. We would exchange experiences about ourrespective gardens, and Emma would often remark how easy it wouldbe if we built a little den down near the woods at Brit hdir. I could still dowoodturning there, and she would no longer have the long drive to bandpractices. As we saw the community beginning to take shape over thenext two years we realised that these suggestions had a lot of sense inthem. it's OK living in an agricultural workers' bungalow in Wales in thesummer, but each winter you become aware of just how badly they aredesigned and how poorly insulated they are. I've lived in four since 1977,and not a year has passed without me thinking about designing a houseof my own that wasn't always cold and draughty; a house that faced andwelcomed in the sun; a house that made sense and was fun to live in;something that didn't cost the earth to build and didn't need a mortgage,so left me free to live simply. )ane and I had 1.5 acres of tightly packedDouglas Fir alongside our track. One day we started thinning the woodWe were going to build a den at Brithdir.


10/121

So many people would love to build their own place, and so few do,especially in Britai n. So many need to live and work and be in nature,yet so few are allowed the space to do so. Ican understand the planners'fears that if everyone were allowed free range to build what they likedwhere they liked this small country would be overrun, but yet there issomething in me that revolts against a system that assumes that Iand naturedon't mix. At the heart ofour planning laws is the unspokenassumpt ion that people and the countryside are bad for each other.This is clearly nonsense - we have only evolved as we have overhundreds of thousands of years, by being tuned in, precisely andacutely, to the seasons and rhythms of nature. lt is as natural for us tobuild an appropriate shelter as it is for badgers. There is no evidencethat it works in the long term to keep humans confined to towns andcities, except for a few mechanised industrial farmers who are supposedto keep us all supplied wi th cheap food. it's not working for wildlife,for the farmers themselves, nor for us, fighting as we have to foroverpriced and bad ly designed flats and houses in fume filled townsand cities. For us to live sustainably on this endangered jewel of aplanet it wi ll take radical re-appraisal of how we relate to and use theland, and I want to be part of the solution, not the problem.

We had to think all these things through before deciding to build aden that would certainly have been den ied permission, had we askedfor it. At the moment (and things could change very quickly) Iconsiderit folly to assume that the powers that be are capable of taking rational,sustainable, holistic decisions. They are too stretched; their offices aretoo noisy; their journey towo rk is too stressful; they are too cutoff fromnature. I worked in that world for seven years. Acounc il ChiefExecutiveonce said to me ul'd love to be green, but I can't afford to. I have tofiddlewhileRomeburns."I don't want to make anyone wrong. We ' re all doing our bes t.I've simply decided that no one has the motivation to sort my worldout more than I have. I have taken reponsibility for my life, so Idecided to build my dream house here, before I get too old to beable to lift a rafter. Anyway, it's in my genes. My grandfather builthis own wooden house. Now I've done it too, it seems the mos tnatural thing in the world


11/121

DesignOn with the nitty gritty. I have real ised, through making plenty ofm istakes, that it pays to think hard before building anything, and towrite down my thoughts as I go along. I have found the principles ofpermaculture (Pc) an essential tool in this design process, and cannotrecommend highly enough that you go on at least one Pc design courseand read at least a couple of Pc books before you consider buildingyollrown home (or even a home for your hens). I started this project by rereading my favouri te book on sel f-build, Ken Kern's Owner Built Homeand his sequel The OwnerBuilt Homestead. Then wrote a simple brief tomyself: "Goal:An autonomous house ofwood; very warm, very dry, cheapto run. Made from pine logs from Erw Deg (where we lived) . Turf/brackenroof... it is built on a slope near woods."'

Over the next year, we agreed the best s ite for the house with Em m a atBrithdir. lt had to be preferably on a south facing slope, invisible from themountain, Carningli, or surrounding hills and near the woods. The placewe chosewas on a bank covered with bracken in one of he smaller fields,just below a pair or hawthorn trees, near the end or a green lane andabout 400 metres fi"om the farmyard and main house of the community.

Brithdir Mawr and Carningli mountain.


12/121

Thewholefarm is t6sacres,rolling fields, big hedgerowswith oak and ash standards,and includes about 6o acresof woods. Big and green andfairly wi ld. So I needed to takesome time thinking about anappropriate bui lding/de n{hoganfcabin for this kind ofla ndscape. Someth ing thatwould not impose anything onthe land or be inorganic andheavy. Something that couldeven rot away when we aredead. Something incorporating Iron ' '' ,dw l'liog"t C>tlllow embodied energy. Lowest building materials for embodied energyare cob, straw and unsawn wood from on site. Also. cement manufacturenow accounts for 10% ofglobal C02 emissions.* OK. No cement. Thiswa s going to be an attempt to build a home whose materials were verynatural and very local. A sustainable home

We had two young people, Dima and Carol, staying at our place as'Wwoofers', working fo r their keep for a coupleofmonths. For one winter(96/7) we spent at least half of the dayl ight hours available thinning ou rbit of Douglas Fir wood across the track. Each day we would note thetrees that were too crowded, too small or that had been blown at anangle by the wind. With an axe and bowsaw Dima and I cut them downand Jane and Carol did the snedding . cutting off all the litt le brancheswith hand a ~ e s In the evenings I read Sheller, the American book onethnic hand built housing the world ove r, and borrowed a book on logcabins. I revisited Castell Hen llys, two miles from here, where there areprecise reconstructions ofa settlement of round wood, mud and thatchhouses from two thousand years ago, rebuilt in their original post holes;a llmadeofnaturalmaterials.The design for this roundhouse started out as an eight sided hoganwith split level roofwhose centre would be supported by enormous forkedbeams. The roofwould be Navajo sty le 'whirling logs' a circular pile of100 trees. The walls would each be 12ft (3-5 ish metres) long. joined at*wurceTomorrow'World


13/121

the corners by overlapped joints. The log cabin book showed how to cuteach log along its length wi th an indentation in its underside so thateach would fit tidily and warmly on the one beneath it. This seems thebest alternative to adzing each one roughly square, or having the wholelot milled by a sawmill, a prohibitively expensive operation. I tried itusing an electricchainsaw. The little machine just couldn't handle thework and I couldn't handle the noise and strain of shaping every s inglewall member along the grain like this. OK, back to the drawing board.Where was that tiny reference to cordwood walls using differentthicknesses of logs, all cut to the same length and stacked up as in afirewood stack (or 'cord') to make a wall? I found a couple of smallreferences. lt was a techn ique used by European (usually Swedish)settlers in the midwest US and Canada in the 19th century, to make'poor people's' houses. Warm, easy to build for a non professional, andusing all sorts ofwood. The book also said that softwood. such as pine,loses heat two and a half times more readily through the end grain aslaterally.To have the equivalent ofa 6"/tsomm thick wall a fai rly standardSwedish style log cabin thickness I would need 2.5 times that ascordwood, i.e. 15" logs. I actually settled for t6",to be on the warm side,and because it translates easily into metric -4ocm. OK it's a weird reason,butit'sthetruth.If you ever get into designing things, you will find that you evolveways ofworking that make sense to you. You have to, because it actuallyinvolves a lot ofconcentration. I have two basic methods. The first is tocarry out a conversationwi th myself inwrit ing and drawings as the designtakes shape. I have 25 30 pages of these pencilled conversationsshowing the gradual evolution of his des ign over about nine months. ltis useful at the end of a design session on, say, the roof, the floor, thewindows, the drainage system whatever to state the conclusions drawnand the things to do next. This is because all parts of a holisticallydesigned building impinge on each other, and often you are left withquestions rather than answers. For example, I le ft any detailedconsideration about what to make the back walls and floor with untilthe JCB had dug the hole into the bank. This was because ifwe had hitrock a foot down the whole design would have needed to change. lt wasworth writing down the questions I was left with after each designsession. Doubts too. I have long harboured dreams ofheating a house


14/121

by the slow convection of ai r from pipes through and around a vastunderground water tank that has been accumulating heat from solarpanels all summer. Fair enough, lets go for it. A5 Iworked on thinki ng itthrough it became obvious that this was actually going to be morecomplicated than I could handle. Maybe you've got plenty ofmoney topay consultants for this kind of thing. Fine, but we had no money at all.This whole house was built on trick les of cash in, using natural andrecycled stuff, and good luck. So if there was a doubt that (o) the waterand plumbing system would be just too complicated and (b) a hugewater tank in the middle of the floor might affect the footing for themain house supports (just a feeling rather than anything definite), Ichose simplicity and went back to the drawing board again .I am sorry if this seems too long winded for you, but I cannot stressstrongly enough that if your house is going to work for you perfectly inthe situation you have, then you will need to take just as long designingit as building it. So much waste in our society comes from people wantinga quick off-thepeg 5olution, and that u5 ually means more transport,less sympathy with the environment, and just silly design. Look at allthose new housing estates with the houses dutifully facing the road,whether the road is to the north or the south. Simply by locating a houseto face the south you are reducing heating bills by at least 13% . And a llthose south facing roofs with not a solar panel or PV t ile to be seen. ltdrives memad. Planning? Don't make me la ugh.The other type of design work is something that maybe you alreadydoverywell,but l'veneverseenitta lkedaboutmuch.Virtuai Realityisnow a concept that many people are fam iliar with, though, so it is certainlyhappening in cyberspace. I hope the Kogi native people of Colombiawill not be offended if I use their phrase 'working in Aluna' to describeit, for that is what I call it. The mind is a wonderful thing. Recentlypsychologists advising an Am erican footbal l team carried out a seriesof ~ p e r i m e n t s on the team members. Success in American footballapparently involves a team in having a series ofmultiple-move strategiesthat the captain can summon up, like a program, in the course of playby yelling out a number. The team then make the appropriate moves tocarry out this strategy. The psychologists worked with the team to gothrough 5everal virtual games, hunched up together in a room withouttouching a real ball, concentrating in their minds, imagining each move


15/121

andthecoun termovesof theo pposition . lt was assumedthat this kind of virtualpracticesessionwouldnotbeas goodastherealthing, butmightjust beofsomeuse.lnfacttheyfound ittobeofmoreuse than a rea l practice, andgot better at it, too. The teamperformanc e dramaticallyimproved. When 1 read this Iremembered the fi lm from theHeart aftheWorld that the BBCmade at the invitation of theKog is, a na t ive nation in RobRoy's cordwoodandmasonryhouse.Colombia that has retained its culture s ince before the conquistadores. AsI remember it, the Mamas, the leaders, do not take any decisions on thephysical plane until the consequences and details have been worked on inAIuna; in darkness, in their minds. Using this idea, Ideveloped for myselfawayof imagining in detail everyth ing about a particular aspect of he housethat Iwas working on. Ifyou do this you wi ll find that you can at first onlyimagine one or two steps or objects, but as time goes on you can developAluna into a very creative space in which questions can be posed, pract icalsolutions put forward in succession, and each possibility looked at in 3Dand full colour. You can turn something around, imagine it in pouring rain,imagine five nine year old boys climbing all over it and so on. As this housebecame a rea lity, I found that at least halfof he lime Iwasworking in Aluna,usuallyon a practical detail that I had never actually encountered in real life.or at least never on the same scale. For me the best time to do this is themiddle of the night, unfortunately, when all is totally still, so Jane got quiteused to me getting up at two or three a.m. to draw some detail of theplumbing system, a window frame or the roof rafters. The Kogi, I think,believe that on the Aluna level a ll minds are linked. I have since found outthat severalideasthat cameto me inAiunahavebeenthesubjectofquitedetailed trial and error work by pioneers in alternative building methodssuch as Rob Roy in the US. who contacted me when my house becameknown about and told me that it was a cordwood masonry house.


16/121

Anyway, there you go. it is a way of designing that soaks in all availabledata and then a llows another part of the mind to mold it. it works forme. 11 is the ma in reason why I couldn't apply for planning permissionin advance, as a matter of fact. How can you put in deta iled drawings ifyou don't know what you are going to do until the night before you doit? From now on there will be a s imple description ofconstruction details.Please don't ask me to explain everything, though, because I've forgottenmost of it. We just woke up the next morning and got on with it. I betthat's how our ancestors worked for the last few hundred thousand orsoyears that we have beenbuild ingsimpleshehers

(diagramrJ This is about QS detailed QS any design got: post Qnd beamstructure, round, with reciproCQfframe roo f Outer circle o fuprights togo up first, then henge crosspieces, then roof Cordwood in walls.


17/121

MaterialsWe took a winter to get the wood cut. Dima, Carol, jane and Iwould plungeinto the damp deep sloping pine woods in the mornings with a sharpbowsaw, two little ~ e s for snigging (I think that's an old word for cutting offminorbranches -the other good word likethat is snedd ing, which is draggingthe big logs out by horse power), a really sharp fe lling axe, some rope, andtough gloves and boots. The inventory I wasworking to was

roof supports (rafters) 6m+ 13 off plus sparespole uprights {9 diam pref) 3m- 13 offinneruprights (8" diam)maybesm-tJOffwalls: a rough guess of210 x 16" (40-cm) logs per full wall, approx24lengths o fabout 4 m each! Say 9 walls max, allowing for windows,so: lengths ofall thicknesses 4 m - 200 off.

So we had plenty to be gettingon with, We got really good at tree felling andt hinning by hand, and measuring by foot. I used a t ipi pole marked out inlengths of4m and 6m with yellow insulation tape, but theworking situationwas temperate forestwith brambles, smalls treams and occasionally almostimpenetrable undergrowth at a slope ofabout 1 in 3 I use metres for stufflike this because 1 metre is one ofmy paces with big Swedish army bootson (vintage 1941! - aren't street markets amazing sometimes?).

Thinning trees for building material in our pine woods.10


18/121

Logs being transported by milk float.I like to have all important building ma terials ready before starting work, somade a call to Kevin Thomas, a friendlyglazierwhose workshop was only 10minutes away by milkfloat. Glaziers get enormous amountsofdouble glazingunits that are perfectly good, but have been taken out in some renovation JOb,or have been ordered in the wrong size for a particular frame. If you can beflexible in your size requirements for windows, as ofcourseyou can with cob,wood, cordwood and strawbale buildings, you save a fortune on windows.One day Kevin called me back. "Got a load ofpatio doors and windows. I'mclearingout theworkshop. I've got to skipa whole load. Do you want tocomeup in the next half hour (sic) to have a l o o k ~ " SCRAMBLE. I bought threedouble patio doors, about three ordinary wood fi-amed small i n ~ . andabout 15 big double glazed units (several too big for one p e ~ o n to lift) fo r[125. just fitted them all on the open back of he mi lkfloat. (Don't ask me allabout that suffice to say that ifyou want to wean y o u ~ e l f off a car, get asecond hand electric milk float. The old batteries will still last you about threeyears. You've got a range ofabout 25 miles for an average 15electricity units,charged overnight, and a float will carry 1.3 metric tonnes.That's a real loadofwood, logs, manure, or windows. Mind you, keeping one going requires asteep learning curve involving DC electrics and battery charging andmaintenance. lfyou're temptedthough, checkoutyour nearest dairythatuses floats, and go for it Make allyour neighbours laugh.)Theonly other materials I bought in advance were 2 rollsofgalvanisedsteel fixing strap from a builders' merchants, and a good supply ofnailsof al l sizes. including 30-40 6"/ lSOmm nails.

11


19/121

EmbOdied EnergyThis is the energy tied up in a structure by virtue of the transport,extraction, processing and other costs that go into all the buildingmaterials. Cement is very high in embodied energy. So is aluminium all that mining, all that e lectricity to make it; transport from, usually,Ca nada. If a ll countries in the world used as much aluminium by2050 as Brita in uses today, the globa l capaci ty to sustain suchproduction would be exceeded by eight and a half times. New glassand bricks (all that baking) are a lso high in embodied energyGenerally, the true cost of road transport, if built into the embodiedenergy costs of bui lding materials . can make a huge impact on theenvironment. Ifyou use wood in London that was felled in Scotland,for example, it will have a higher embodied energy than the samekind ofwood imported from Latvia, because the latter comes by boatto London docks, whereas the Scottish wood has come maybe 6oomiles by road. This is one of the critical, yet almost invisible, factorsthat form part of the big issue of liv ing sustainably in everything wedo. Seeing just how much wasted energy goes into the building ofan ordinary modern house was one of the main reasons that droveme to designing and building an alternative model from scratch. Ilearned most of it by joining the Alternative Technology Associationrun from CAT(Cen tre for Alternative Technology), the Ecological DesignAssociation and reading books like Our Ecological Footprint and theFOE publication Tomo rrow's World (see Appendix AJj).

The house wi th rOCk bottom embodied energy, and thereforethe kindest to the ea rth to bui ld, would be made of natura lmaterials found and processed by hand on site. Ou r ancestorsbuilt houses from cob (clay with some sand and straw), localstone, wood and thatch. Nowadays I would con tend that by addingstraw bales and a modern technological miracle large seamlessrubber pond line r s hee ts to this list (adm ittedly rubber isimported , so higher in embodied e nergy, but grown in sustainablecondit ions giving livelihood to rainforest dwellers), we can stil lproduce warm, dry and organic houses at a fraction of the cost toourselves and to the environment.

We brought the wood the 19 miles from our old place to this intwo big lorry loads, the lorry being specially equipped with a longlZ


20/121

This is the amount o fwood needed or the structural fromeworlc.lifting arm and chains. We then carried it the 400 metres down thetrack by tractor and trailer; maybe twenty loads, spread over the nextthree months or so. We used no power tools on site, so I cut up thewall log-ends up by the fa rm house with an electric chainsaw. (I'llfollow American custom and call them log-ends from now on). Soeven a low impact building like th is has quite a bit ofembodied energyElectricity was consumed for the cutting. Some was renewable ene rgy,mind you, because our community bought an inverter that enablesmost electric appliances such as hand tools, electric chainsaw,computers etc. to be run from the battery bank fed from streamturbine, solar PV panels and wind turbine. Diesel was used for maybe80 mi les of lorry transport pl us maybe 5 miles of tractor use. I havesome regrets that even the small embodied energy costs in this housecou ld not be avoided, but it is pointless being aga inst all technology,isn't it? We didn't have to build an earth sheltered house into a bank,but I'm glad we did (there's a great gale blowing around outside as Iwrite this and we hardly not ice it in here) and I'm grateful a JCB wason hand to save us weeks of back-breaking digging into day.

The jCB revealed the perfect material for filling between the log-ends-cob, as we had hoped. Ifwe had hit rock a foot down I would have hadto redesign the whole thing, so flexibility was crucial. We made three

13


21/121

Inside the bender (inset} the outside view.piles subsoil to ei ther side of the hole and one big pile of topsoil forspreading back around the house and for the roof.

By September 1997 we had the main logs by a big clayey hole in abank. jane and t set up a bender at the top of the bank, made of benthazel poles covered by canvas, and borrowed a little wood burningstove. This was our home for four months. Don't attempt this withsomeone you don't know very well. At times, after weeks of rain,mud and sleet. it felt like Scott of the Antarctic's last camp. Slugseat ing our books; wheelbarrowing batteries through ankledeep mud;wet clothes, aching bod ies. t won't go on any more about it, butplease don 't think that build ing your own place is some kind ofeasyoption! Right. We're ready to start.


22/121

The SkeletonA traditiona l and safe way of building a simple house is to make a 'postand beam' skeleton and fill in the spaces. The skeletonof his roundhouseis a wood henge of uprights with cross pieces forming a complete circle,and a reciprocal frame roof, all of round Douglas Fir poles.The reciprocalframemeans that the raftersrestoneachother inacircleatthe centre,where there is a hole. More strength , for a turf roof, is supplied by aninner circle of supports, but the outer circle comes first, and the size oftha t was actually dictated bythe A t spacewe had availablewhen Berwynleft onhis JCB.

Firstwe laid out the rafters on the ground to get an idea o f he thing.it's no use build ing something to some theoretical size i f n practice theends ofyour rafters are a bit weak and dodgy looking. it' s always worthhaving something small and strong rather than maximum width andsagging in the middle. SoWillow (a permaculture friend and t ree planterextraordinaire) and I laid out the strongest of he poles around the circle."How many poles you having?" asks Willow. "Twelve ," say I, becauseI've already made one wi th twelve poles in the field across the way, thatwe still use for a marquee frame. "No," says Willow, " it's got to bethirteen. I'm not going to help you build a housewith twelve sides. it'sgot to have thirteen." it's no point asking him why. Willow can talk allday about the moon phases and the Mayan calendar and stone circlesand just about everything. In a quick menta l resume of the last ninemonths ofdesign work I thin k, "How attached am I to it having twelvesides?" The answer comes back - not at all. The more s ides, in fact,within reason, the better, as each cross piece hold ing the sub-raftershas less weight to hold. "OK, thirteen," I say, and from that moment onall thought of attempting to calculate measurements in advance goesout the window. Try d ividing anything by, or multiplying it by, thirteen.From then on it was rule of thumb. just as well really, because I didn'thave a tapemeasure long enough to reach across thewhole house circle,and still don 't. The circle is not exact it fits the space and suits thestrengt h of the rafters, but I still can't tell you precisely how wide it is.

We laid out the thirteen rafters and stuck a stick in the ground whereeach support would be. Then we started digging holes and painting

15


23/121

creosote on the bottom metre of the main supports. I know that polesrot if you put them in the ground, especially in the area just aroundground level, where bacteria and fungi have moist air and damp woodto work on. Oak and chestnut resist rot for 40-50 years; sycamore, a ~ e land ash wi ll last about 18 months, and most woods are somewhere inbetween. On previous projects I have tried charring the bottoms as ourancestors did , creosoting, standing the pole in a plastic bag with creosotein and putting concrete around the bag, and doing nothing. After aboutte n years, none have rotted away yet , but I guess the 'doing nothing' wil lgo first. If we were using concrete here, we could stand the poles onconcrete plinths, "steadied by stainless steel pins, but then the uprightswould need much more bracing than they've got, to stop them twistingor moving sideways. Call me old fashioned or unrealistic but I don't th inkyou can beat a nice hole with the pole stuck good and fast in it. This wholebuildingwill rot anyway, one day. I don't want the poles to rot faster thananything else, though, so each time I clean out the chimney I put thenatural creosote and soot into the clay around the base ofa different pole.The holes area cubit deep. Here's jane f1nishing one off. When she cantscrape any more from the bottom of the hole, that's a cubit.

jane digs a post hole; she can j ust reach a cubit down.16


24/121

When all the poles were in,wecutthemsothatthetopswere all exactly leveL This isimportant so that the wholeplacedoesn't lookcompletelyskewwiff. The load from theroofisdistributedevenly, andI could use the henge crosspiecesas a levelwhen itcameto fitting windows, etc. Howto do it is to make a bunyip.Not a word you see much, isit? A bunyip is a length ofhose of approx tom lengthwith about a metre of clear The posts go into position.plastic at each end. If you fill this with water, making sure that all airbubbles have left, you will find that the level of the water visible in theclear plastic ends is the same, no matter how far apart you hold the ends.Don't attempt this wi thout a friend helping you. You nearlygo mad as itis, because it's all too easy to move one end up or down too suddenly,with the result that a jet ofwater comes from the lower end, and you haveto replace the wa ter and start again. Still, when you get the knack of it,which can involve marked sticks attached to each end, pieces ofcha lk, communications ski lls and a bit of cursing, you end up with aseries of marks on the uprights that you can guarantee, by scrupulouscross checking, are all level with each other. I started with the pole in thesouth west that I knew I wanted for a door post, and made a mark zmfrom the ground level. (The other time you may find a bunyip useful willbe ifyou want to dig a swale orditch fordra iningfwater retention along acontour.ln thiscaseamarkedstickagainsttheclearplasticisessential.).

At this point we have 13 ma rked uprights in holes. I actually cut thetops off as they were standing, as I could not manipulate them singlehanded (each pole was about three metres long and at least zscm indiameter, so weighed more than me). If there is a team of you, you cantakeeachpoleout,cutittosizethenreplaceit, but l findearthfallsinthehole and I can't be sure of the level again, so I cut them where they are.

17


25/121

Each pole stands on a rock at the bottom of the hole, is jammedtight with stones, then tamped upright using a plumb line. Don'ttry and put up your main supports without test ing that they areplumb straight. No point tempting gravity, is there? There is a casefor drilling the hole in the top of each upright before you do thefinal tamping, because the pole moves a bit as you lean against itwith a step ladder. Drilling these holes , 3 4 (75100 mm) into theend grain zm up a step ladder, with a brace and bit, wa s the singlemost exhausting operation of this entire project, by the way, so whenyo u've done that , relax. The rest is downhill.To make the henge, I measu red the exact dista nce between thecentres of each upright at ground level, and cut the cross piece tothis length plus one pole diameter. This is importa nt so that yourhenge pieces overlap and are pinned together on the top of eachupright (see diagrams 2 and J). This is a way of doing it that I haveevolved over ten years or so . If you ca n do it better, by all means doso. it is important to use the henge to hold itself as a strong circle,to al low for most of the roof weight to come down each suppVrtevenly. The pin going from the up right through the cross piecesinto the rafter ensures structural strength, and acts as a usefulfulcrum when adj ust ing the rafters. At first it is a fiddle cuttingeach cross piece to get the angles reasonably tight, but actually itdoesn't take that long, and looks great. As the electric chainsawwould save a lot of lime wi th these overlap joints, I cut the crosspieces in kit form up by t he co m muni t y workshop, usingmeasurements of each gap marked out on a piece of paper. I havefond memories of this scruffy, damp piece of paper being passedfrom jacket to trouser pocket and back with the circle of postsmarked out and 223, 219, 226 etc. in all the gaps. (I find ems arequite good for measuring this kind of size, then feet fo r a bit bigger,then metres for long lengths. Wh at do you estimate and measurelengths in? Be honest . it' s a mess, isn't it?)I had the henge fin ished in two weeks, working every day, usuallywith Emma as helper. to whom I am eternally gratefu l. At th is pointI could see what the size would be the henge conta ins the space,and the imagination starts filling in plenty more ideas.

18


26/121

3-.... 5 1 ~ 1 -.. J::._ ' " ' " ' ~ /" c ' ~ ~ l ' " ~ _ .~ : ~ ~ :: ':t;: ~ ~ ~ ' : ' : : ; ; . ""-. t . ..n""' " v . ... # ..H,.,f (-" '"::1 $f>"Af t ; . , ; l ! , . t ~ ,.{/o-.1' ,..._.f_s , ... f"-, c..OSJf'I;."",S

(diagram 2} Henge, crosspiece, rafterfixing detail.H ( . . , ~ o ~ . < . f U ~ . S oote.rff 4.., .. ( < . 'c. c ot" .._r., .. " " ' ( ! ' ~ t f ~ r , , H-.( rri-Jn=F~ ~ . 1 - c . .

(diagram J) Arrangement and oint detail for henge crosspieces.19


27/121

Reciprocal Frame RoofEver since build ing my first reciprocal frame roof, on a simple buildingat our last place that we called the Dojo, I have been fascinated by theease and s trength of them. The idea was made popular by anarchitectural firm called Out ofNowhere, but Idon't think they inventedit. Our friend Jack Everett built his do jo near Stroud with one in theea rly 198o's. An yway, it is a great way to use round timbers to hold aroof up without need ing a central pi lla r; you get a hole for a skylightinstead. The essence of a reciprocal frame is that the rafters a re selfsupporting at t he centre; each resting on the one under it, in a circle.You do not attempt to hold the rafters at any part icu lar pitch you letthem rest on each other and the pitch is what you get. The weight istransferred evenly down each rafter to the up rights , and vertically downto the ground. Putting up a reciprocal frame is great fu n, but has anelement ofdanger too. so it's best to ensure that there a re no childrenrunning around when you do it, and that if you have helpers, they arenot afraid of long poles rolling about above their heads. (Some peopleare - surpris ing, isn't it?)Now we have a henge ofposts and crosspieces, with a hardwood pegprojecting about 20Cm above the joint. The essential extra for erecting areciprocal frame is the Charlie stick - a strong pole in aY shape strong

Dojo at Erw Deg.zo

enough to hold all the raftersup until the las t one fits inabove the second to last andunder the fi rst one, when theCharliestickcan be removed.I have a rule of thumb withthese rooFs that seems towork:the length of the Chorlie stickmusl be: henge height + (no . ofroftusx lheira verage thicknessat their ends) + 1 footjJO ems,minimum. For this house theCharlie needed to be at least 14ft; that's about 45metres long.Such poles are not easy to


28/121

The netlrly completed henge, with t l crosspiece t ~ w o t i n g f i t t i n gcome by, so when you find one, mark it, tell all your helpers about it, anddon't cut it up. From the 250 or so trees we felled I found one, and onlyone, perfect Charlie stick So now, a year later, its time had come. First Ibuilt a stepladder tripod like an adventure playground climbing frameabout4metreshighinthemiddleofthec ircle,usingthe longeststepladder we had. This actually felt much safer to climb on than the stepladder itself, as the base was well puddled clay, and the step ladder wi thitsneat littlemetalfeetwouldgraduallyti lt toonesideuntil itfellover.

Next I set up the Charlie st ick in its place. This needs some thought.The best way to get an even circle holejskylight in the middle of yourreciprocal frame is to have a standard a lignment ofeach rafter pole. Ifyou want a clockwise spiraljiris effect when yo1.1 are standing under yourroof, as Id id, then you need to move clockwise round the circle in placingyour poles. and each pole has to be displaced a standard amount to theright of the centre of the circle. If you are using an even number ofhenge uprights, this is a matter of standing at one upright, looking atthe pole directly opposite you, then aiming at, say, the centre of he space

21


29/121

to the right. A point near the next upright wi ll give you, ultimately, a largecentral hole - a point near the centre line will give you a smaller hole. Inthis case, as we had an uneven number of uprights, I aligned each postnear the post on the right of the space opposite (diogrom 4). This meansthat the Charlie stick, which has to hold the fi rst pole, must itself be offcentre. l wanted a hole about a metre across, so figured that ifl offset theCharlie by half a metre to the right, that would work, and so it did . TheCharlie was put up standing vertically on the ground, gripped in the jawsofa workmate to stabilise it lower down, and lashed to the wooden tripodto stabilise it high up. lt is good if you can also have an assistant to helpyou to hold the Charlie stick as the first rafters are being put into place,becauseifitcollapseshalfway throughit'slikePickupStickswith hundredweight heavy, 6 or7metre long sticks. Igot all the best poles ready, fanningout from their uprights, and drilled a 28mm hole in each aOOut 3"/7smmdeep, at a slight angle, and allowing at least half a metre or more for theeaves. I also made a slight notch in what would be the undersideofth'!rafters to allow a good sit on the henge (diagram 5) .

lt is at this point that good pole selection pays off; each pole must bethickenough at the centre to be capable ofholding a considerable weightatits thinnestpart, whereit meetstheo thersatthecentre. lfapoleistoo long, you can always cut off the surplus at either end, so in thisprocess it is always best to use good st rong poles that are too long. Thesurplus lengthat theeavesis actuallyquiteuseful,asyoucanma nipulatea pole much easier around t he fulcrum of he henge ifyou have 2 metresor more of heavy end to play with. Several of my poles had this muchspare, so I d rilled the pin hole accordingly further from the end . Theoutstanding ends were again useful, when we clad the roof, to clamberup on to the roofa nd to lean secondary rafters against. The houses builtby our ancestors from cob and thatch, such as the ones at Castell Henllys,often had huge spacious eaves coming down almost to the groundThese would provide drycover for animals, firewood, hay etc. and wouldof course protect the outside wall from the elements. These days, longrafters would also provide the basis for greenhouse extensions in thesunny sectors. Between you and me, I have often found our generouseaves alsoofuseforkeepingdrywhile peeing inthe rainRight, so our rafters are d rilled. and ready. The Charl ie stick is

zz


30/121

I , ~ : ' ; : r : t ; : ,, I , , ..~ r , . ~ r ~ ..

If : r ~ A ~ ~ '"7""'. ~ ~ ~ ~ . , , . f: ~ r ; v ~ ~/:> 1;-l.._ - < ; ~ t .

l ~ . o ; f l ! p ~ / . - : ./. , a . : ' i i . ~ , . t :

(d1agram 4) ReCiprrca/ roof rame rafter layout and plac:mg sequence.

w,fl- ra{-fu "f5-'/' ,; . . , ~ ~ .. K" J""""..


31/121

waiting. I had a spare, smaller Charlie of about 2.5 metres long (let'scall it Ajusta) for adjusting the rafters once they were up. So I lifted upthe first rafter (in the South; no reason, but Willow would probablyapprove), heaved the thin end up onto the henge, climbed the tripodholding the end of the pole, and, on tiptoe, lifted the first pole up andinto the Charlie Y. Iwas not actually being assisted that day, so that tenminutes or so was the nearest Ihave ever got to being a circus performer.I got a bi t of birdsong for applause, and tied the pole loosely, to allowsome movement, with baler twine. (Baler twine? We use it as currencyhere. I once assumed it ' s minted somewhere and planted by God inhedgerows and on o ld fa rm machinery the world over, but find that youcan buy it in big rolls from agricultural Co-ops.) I then returned to theground, lifted the heavy end until the hole was over the peg in the support,and let it down. With each junction at the henge, Iuse a pieceofbuilders'galvanised steel f i ~ i n g strap, which comes in 5 metre lengths. I cut offabout a metre, and half it with a couple of nails on the upright, thenone nail on the top of he rafter. That stops the rafter rolling anywhere,which is what they sometimes try to do. Stop to take a photograph.Note the pegs sticking up from the henge, the Charlie stick and thetripod


32/121

From here it is logical andstraight fofWard, going roundthe cirde adding each rafter,aligningeachtotheappropriateplaceacrossthecircle,adjustingeach one at the centre withAjusta, tyingwithbalertwineandhalffixingwithfixingstrap.

If you are likdy to bebuildingoneoftheseyourself,may I recommend that you dothis process with a scalemodelofsticks flrst,saytenthoronefifthscale. There is nosubstitute for eKperience in 1996Jield marquee slructure.making the small adjustments of each rafter as it goes on so that thecentral hole becomes regular, and it is much better to get the hang of itwith, say, lm sticks than with giant ]m ones. Having said that, I stillbelieve this system is pe rfect for people with virtually no buildingexperience,solongasyouhaveplayedwith long poles, levers etc. a bitNow we come to the last two stages of the reciprocal frame skeleton,which are The last Pole and Taking Charlie Out.

The last pole fits under the first and over the second to last. lfCharlieis too short there wi ll be no room for the last few poles. Ifhe's too longthe poles will keep trying to roll away down each other and Taking CharlieOutwill be a bit dramatic. With Charlie at formula length Ihad just enoughspace to squeeze the last pole in to its space and lower it down onto thepeg with a bitofgrunting and groaning. lt is now just a matter of removingCharlie. Of course Charlie, by now, has come to realise that he, and healone, has been holding up about half a ton of woodwork, and isindispensable to the entire project. He is well rooted in thick clay. Removinghim feels hazardous. even if it is not. I will always remember when wewere at this point putting up our field marquee structure in 1996. One ofthe observers, wringing he r hands in dismay at the apparent certainty ofimminent loss of ife, wailed, "What does it say in the book?" Well, here'sthe book, four years late, and it says Don't Panic, take it slowly, and dig

25


33/121

down under the Charlie gently unti l, bit by bit, the structure takes theweight, the poles settle, and Charlie can be removed, preferably to carryout a ceremonial function somewhere. I adjust the circle wi th Ajusta,then fin ish pinning down all the fixing straps across the tops of therafters and down to the top of he upright on the other side. I then fixedthe poles at the centre with short lengths offixing strap on the top sidesof the poles, and by my favourite fixing system a metre of telephonewire wrapped round several times and stapled two or three t imes. Thebasic skeleton is now complete.

Detail o f elephone wire strapping around the skylight hole.26


34/121

Turf roofs are heavier thanalmost any other kind of roof.Ifyour house is to be built in aregion where winters bringheavy snowfalls, you need toestimate how m uch strengthyou need in your roof to holdup not just athousand or morewet tu rves , but also a layer ofsnow. Heavy snow fallsa every

The Full Skeleton

rare here in this part of West The roofwith a typical light snowlood.Wa les, so this roof is not built for the worst case scenario. I also did notintend to use turves much thicker than about 3-5 inches, or 9cm, as itrains often enough in the summer to prevent the turf totally drying outStill, a circular roof 12m/4oft across gives me approx 1400 turves ofonesquare foot. Each turf weighs from 1kilo (2.2lbs) to a max of 10 kilos,averaging about 5 kilos, so for this roof we are ta lking about somethinglike 7 tons of urfwhen wet, plus insulation ofabout a ton. Looking againat accounts ofethnic round houses, I noted that usually there were innerposts to bear the extra weight ofa thick roof. The circular earth lodges ofthe early British, and No rth American Native peoples, the Mandan, thePomo and the Miwok, a ll used an inner henge of strong poles to supportheavy roofs. This house therefore also has an inner he nge of stronguprights and crosspieces to support the load-bearing rafters nea r theirmid point. If you are worried about load bearing capacity of roofs, thereare experts who can calculate these things, so consult one. My basic ruleof thumb with this house was: (a) to bu ild with the strongest members Icould actually lift, because in this siteand weather it would be si lly to havealways to rely on one or more extra helpers all the time; (b) to useexperience of build ing wi th th is type of roof on seven or eight previousstructures; and (c) to trust the designs of our ancestors who built andrefined roundwood and turf designs over thousands of years . Overall,this whole structure feels strong. but in a soft way, as a visitor once put it.As in a basket, there's a bit ofgive in everything; it's not completely rigid

27


35/121

Inner UprightsThere are thirteen inner up rights, each supporting one of the mainrafters. For the exact location of each, jane and I walked around theinside of he structure we had erected, and imagined how much spacewould be required for each sector. As we felt OK about the space forthat, we put a stick in the ground and moved on round. We wentround clockwise talking it through: "Here we are in the coats area;we come in the door, kick offour boots, hang a wet coat on the post ..here. Now we are in the clothes storage area- two racks of clotheshang from the wall to a low partition .. here." etc. We wanted aninner room big enough to dance in. and have band practices ormeetings. so we tried to keep the outer ring reasonably narrow. Inpractice this worked out at the inner circle being about 2m in fromthe outer one. widening a bit in the north and narrowing a bit in thesouth sun bitfbath area. I've never actually measured it a ll the wayround. Hard to believe, isn't it, but I didn't need to.

The way we built the inner henge was to put each upright in first,then fit the crosspieces in situ. Kit form wou ld have been possible,but this way each piece is custom made to its own rafters -less chancefor the working of the great Mu rphy's Law ("Yea, ifsometh ing can gowrong, it will"). Going round to each rafter in turn with a ladder, Imoved a plumb line up or down the rafter until the lead was at itsclosest to our stick marks. )ane then put a new marker stick to markthe place exactly; I marked the rafter where the new upright wouldmeet it, measured how long the support would be, marking it downon a piece of paper, and we moved clockwise to the next. When wehadfin ished,Janestarteddigging theho lesacubitdeep,and l starteddragging the poles into position, their bottom ends already debarkedand creosoted. These polesaretheheaviestinthe s tructure, becauseof their length, and they are all at least a hand span thick, so eachone must weigh at least 100 kilos. Icut them to length with an angledsaw cut (diagram 6) to give a better fit at the rafter. and also debarkedand flattened the rafter at this point. Ifyou want to be ethnic and usea good scrabble word here, pick up your adze. If you don' t have anadze, you can do this kind of )Ob quite adequately with a hand axetapped like a ch isel with a hammer or mallet.

28


36/121

Nof : p ~ l x: ;, , . ; , / f ' ~ f / " f ' ~ / , . . f w ~ n e . 4 !Nf r -hraf-iu . Y o ~ s f.vt ~ t f ~ t . r ~ . . , t.. 1-.; t.. , . ~ . l ~ . - t - ~ . . . , . . ,

~ f"L, .s p ~ r . Tr;J . ... ~ f""" .J('-ro. olf


37/121

CrosspiecesCrosspieces are only about 1m - zm long, so will not be too heavy inhardwood. I used a few strong looking pieces ofdouglas fir, then piecesabout upper arm thickness of ash, hazel and thornwood. (Any woodknown for its strength would do. Oak would be good.) For each piece, Imeasured the distance between uprights where they met the rafters ,and cut an inset into each upright to make a simple joint (diagram 8). ltis quite a tricky one, because of the angles in three planes, and I gotnoticeably better at it as Iwent through all thirteen. You have to seat theleft side of the cross-piece as fi rmly as possible, resting on the top ofthe upright. Flattening the s ide of the rafter at this point allows 2530mm for a ledge for it to grip. (The upright is th icker than the rafter. sothere is a bit to spare.) In places, Igot the cross-piece to overlap or reston its neighbour, but this depended on their size, and the overall 3Dconfrguration.lt is obviously critical not to weaken the rafter here. I predrilled nail holes , then thumped the cross-pieces home with a ma lletand used two longish threaded nai ls for a good grip. Fina lly the fixingstraps are wrapped over the cross pieces and fastened down, linking asmany pieces as possible.

J,: . ; " J o/ , , . ~ . , f ' ~ " ~ fo ; . , . , tr "f"'jl..( ss ~ ~ ~ ... . t ~ ' ' ~ l h ' . 5 . c t r ~ ~ ~ ' " l t l l t l .. 1 t t . ~ f "'" r:v, .."""' fio- t-1-.(. ,..,$,;,;( fock , .J ov f :

30


38/121

BracingI have found the advice of other people invaluable at certain points intheconstructionofth isroundhouse.Someth inglikethis reallygetstheirimagination going, and I had some really useful tips from engineers,builders, architects and craftspeople. The bracing is thanks to aneighbour, John Hargreaves of St. Dogmaels, who pointed out to methat although the structure by now felt pretty solid, vertical poles withcrosspieces could still all twist sideways under the stress of, say, a treefalling on the house. Several other people said no, it's fine as it is , but Icould see his point. Diagonal bracing makes a huge difference to therigidity ofuprights, and it is better to be safe than sorry. So the six backwallshaveabigthickdiagonalgoing from corner to corner.fo rming a kind ofherring bonepatternaroundhalf thed rcle.Each diagonal was offered tothespace, and cuttofitintoanot ch cut into the uprights,thumped home by mallet, andfi xed with 6" nails. l t didn'ttake too long to do this -maybe a day- but after this the Diagonal bracing on the rear arc:.st ructure felt solid as a rock. The drawback is that it is much more difficultto infill a triangular space than a square space, so these wall sectorswere fiddly when it came to filling in the walls.

The structure was now well defined, so felt good to be in and look outfrom. As I have mentioned earlier, we have on this community a simplerversion, without the inner circle and bracing, that we use fo r gatheringsand summer dances, parties, weddings etc. Brent made a grand canvasroof that can be suspended from the rafters and tied to the uprights, andwe have canvas sides that can be laced together. A circle is themost efficientshape in terms of the space enclosed in rela tion to the structure materialused (Rob Roy, in his book, CordwoodMasonry H o u ~ b u i l d i n g . works out thatit is 43% more efficient than a rectangle, for example), so this circular frameconstruction is the best for theTardis effect. You can frt a lot more in than youwould think, and the space looks smaller on the outside than the inside.

31


39/121

32


40/121

The RoofHere is how the Miwok, natives ofwhat is now northern California, builttheir assembly house roof:

"A thatch ofbrush, topped with d igger o r western yellow pine needles,never sugar pine needles, was next put on. This was followed by the finalcovering ofearth. Altogether the roofwas 1.5 or 2 feet thick. The openingin the top of he conical roofserved as the smoke hole, the fire being built

d i y under it. The entrance was on any side. Certain niceties appear inplacing brush and earth on the roof. The first layer of brush, which was la idradially over the numerous horizontal roof timbers, was ofwillow. On thisanother layer at right angles was placed. The third layerwas ofa shrub withmany dose parallel twigs that kept the earth covering flom leaking throughand resisted rot. The proper depth of he earth layer was 4 or 5 inches andwas measured by thrusting in the hand. The proper depth came to the baseof the thumb.H(Shelter, quoting from The Californian Indians).

As far as I know, this roof is unique, and so far it has worked reallywell in keeping in heat, in letting in light, in providing habitats andgrowing places for countless plants and creatures, and in looking greatfrom inside and o utside. Working in Aluna after read ing about roofs, itoccurred to me to use the basic materials that the Miwok used, but tohave a single waterproof membrane instead of t he waterproof sh rubs,and to use straw bales as insulation instead of pine needles. The s trawbales could be tied together, so adding strength as a kind ofskep sittingon the rafters. (A skep is not on ly another great scrabble word but theold type of beehive woven from straw.) For the membrane I phoned upseveral pond liner makers for samples. PVC is cheaper than rubbe r, butits manufacture is grossly polluting. (Greenpeace d id well to persuadethe government not to use PVC for the Millenium Dome for just thisreason.) Rubber extraction is one of the few ways that native dwellers inthe tropical rainforests can make a living sustainably, while preservingthe ba la nce of the forest, so we determined to use rubber for themembrane. This is the single most expensive item in the build ing, costingabout [650 for a 40ft square{147sq.m sheet of1mm thick rubber with aguarantee of 30 years' life. When it arrived (from Stephens Plastics,Wiltshire, tel.01225810324/S/6), it was in a huge bundle weighing over 300

33


41/121

kilos nearly a third ofa ton! We would need some help manipulating thisThe roofdesign incorporates tw'o kinds oflabour input. The first is the

usual one of me, )ane and maybe another helper pootling away at somelong task that can nevertheless be done by one person. This would all be inpreparation for the second mode- one dry day ofant-like activity by maybe20 people, carrying and humping up bales of straw, tying them together,and then covering thewhole thing with the enormous pond liner.After that ,with the structure effectively under a giant umbrella, we could return tomode one, and carry o n bu ilding the walls and turfing the roof in ones andtwos. (This is actually a main reason for choosing a 'post and beam' way ofbuilding, which this is . You may remember that such a system, i.e. roofbefore walls, was proposed by Noddy in Noddy Builds a House , by EnidBlyton, but laughed out of court by Big Ears. The advantage is that, aftermaking your skeleton, you can put your roof on and then work in the dry.Makes a bigd ifference, especially with a Welsh climate. Post and beam alsooffers a bener resistance to earthquake damage.)

)ane, julian and Iare members ofa local permaculture group that forseveral years met on the th ird Sunday ofeach month at different people'shouses to carry out somework, usually ofa heavy gardening/landscapi ngnature , that would have been very onerous for that person to doindividually. Over the years we cleared a lot of bracken and brambles,planted thousands of rees, and dug several fine ponds. We booked ourgroup to come hereon Sunday, Oct 19th 1997, and prayed for sun . Hereis the roof in cross-section (diagram 9)


42/121

Secondary RattersForeachraftertherearesevenoreightsecondaryraftersparallel with itand spaced evenly about a hand span apart, resting with their light endson the next main rafter, and with their heavy end resting on the outerhenge. Secondary rafters therefore vary in si:te from 6m or so (the onenearest the main rafter) to approx 2m nearest the next rafter (diagram 10).Putting these up was easy. Iwas glad that we had thinned so many trees,and therefore had a good choice of over 100 poles to use. I fixed bothends with a double loop of telephone wire, stapled to the pole and to thehenge or rafter. Telephone wire is great if you can get it, Decause it'sdesigned to wi thstand enormous tension over the distance betweentelegraph poles. I was given about a quarter of a m ile of the stuff by BTengineers after the January storms of 1990 brought down most of thetelephonelinesnearus,andhavebeenusingitonstructureseversince.Heaven knows what I'd have used otherwise. Maybe I'd have taken upstone masonry instead. Anyway, if you are bu ilding something like thisyou're going to have to ~ e r c i s e your imagination as to what to use. Fixingstrapisfine,ifugly.butwouldbeveryexpensiveinthequantitiesneeded.(At least 100 metres) Old electric fencing isn't bad, and old wiring is OK.Look in skips outside banks being refurbished. I found about wm oftelephone wire being thrown away outside Lloyds Bank in Ca rdigan.Windows too. lt's amazingwhatbankschuckout.When all the seconda ryrafters were on, we could seethe pat tern of the roofunderneath for the first timeI'd only ever seen it before inAluna. I'm very pleased with it-it'slikeagiant iris ofaneye.ltbecameclear atthis pointthattheinternaldesignshould notimpede the clean sweep of helines of the rafters, so thatmeant new thoughts about L_________ jpartitions, etc. (diagram 10) Secondary raf ter layout.

35


43/121

Willow LateralsThe next task was to weave and fix on the next lateral layer, which ismainlywillow wands and saplings ofaboutthumb thickness. Thesewerecut green, all from within 100m of the house, and used immediately.The best withies were from a coppiced willow, or where a large willowtree has fallen over but continued growing, as they always do, by sendingup dozens ofshoots along the now horizontal trunk. Shoots ofcoppicedhazel are good, too.The aim of these two layers, ofcourse, is to spread the weight of thetu rf as evenly as possible over the entire roof, so I wanted lengths ofwillow no more than a hand span apart over the entire roof. In somecases we managed to weave the strands in between the rafte rs, but wefound that tying with string or nailing was more practical. This task tooka lot of ime to finish, and was a ra ther uncomfortable working position,although theviewwas fabulous.

jane willowing the roof36


44/121

A light coloured tarpaulin covers the willawedframework.Materials

Before the Permaculture Roof Day I had to make sure that everythingnecessary would De to hand, so we brought down the membrane on thetractor trailer, amassed a fortune in lengths of baler twine, and boughtand collected by Land Rover and trailer 150 bales ofstraw.These cost only[100. Strawmust surely be the most underrated insulating and build ingmaterial in Britain. Another important layer in the roofis thecanvas, whichwas to be tied onto the willow layer. Its purpose is to prevent bits ofstrawfalling down into the house, and generally to contain the straw layer. Iasked Willow for any information he had on old tarpaulins that might besuitable, and he found for us two small tarps and one giantwhite pieceofcanvas that had once been the side of a circus marquee. The whitenesswas good, too, as a light ceiling helps brighten a room.

37


45/121

RoofRaisingThe big day dawned unseasonally warm and sunny. Twenty friends andrelatives rolled up, with food to share for lunch. Julian and two twelveyear olds started wheelbarrowing straw bales from the stack to the eaves.Philippa formed a team tying on the canvasses.

When the canvaswas on,Willow and Emma formed a chain heaving thebales onto the roof, and others JOined them. Istarted with a square ofbalesround the roof hole, and worked out in a spiral. Keith tied each new ba le tothe one behind, and I tied it to the one or two above. Children helped stuffany gaps with loose straw, and got into a whole social scene on top of thehouse as the giant skep grew in size and the day progressed. Baler twinebecame as rare as gold dust, but just lasted out. We stopped about 1mfrom the edge, to allow for a slope down to the eaves (no point putting3ocm ofinsulation on the roofoutside the outer wall, is there?) There were30 bales left, so we had used 120 . That's a lot of insulation.

Meanwhile a team ofsappers had been constructing a wooden bridgefrom the bank behind the house onto the roof. They then unfolded thehuge membrane and rolled it up into a 40ft long sausage. As the lightwas fading and Keith crotcheted his last bale link, we all lifted a bit ofthe sausage and marched over the bridge in a long line onto the roof,put it down and unrolled it. The straw was safe, the cover was on, (andthe roofcould hold twenty people). We went for a swim in the stream towash the dust off. I think this was one of the best days of my life

38


46/121

39


47/121

> ' ..J ......Measuring up for the eave logs.

40


48/121

Eaves(RoofedgeThe policy we adopted from this point was to work on the eaves and,subsequently, turfing, when the weather was dry, and todo the cordwoodwalls when it was raining. Here is what we d id to finish the roof.Firstofa li i went round with a sharp knife and cut offthe four cornershanging down, and trimmed the whole sheet to about a foot (3ocm) allround. The spare rubber was useful in dozens ofways primarily to putat the baseofeachwall as a moisture barrier.

A visit ing architect recently looked at the rainwater dripping roundour eaves at various points and sighed 'Oh, if only Icould get away withdetails like that!'. What he meant, in the nicest possible wa y, was thatthese eaves are someof he most unprofessional, untogether. and. well,organicfeaturesofthewhole place. I explained that, rather than try tochannel the water running off the eaves in to a system ofgutters, whichwould be a means of saving ra inwater and sparing the rafter ends fromrot, I had simply put a system of turf-retaining logs that would be easyto replace when they rotted away. This is because the pattern of therafters coming out over the henge is real ly irregular. lt would have beenpossible to make a system of runoff points around the circumferencewhere water spouted off in a predictable fashion, but I really like thenon-linear feature of it just as it is. Ifwe did not have a free source ofexcellent, d ean water from the mountain, I would a lso have put moreenergyintowaterconservation,as lhavedoneonstructuresinthepast .jane and I lived in a bungalow served by a water meter before movinghere, sowere able to see that by employ ing water conservation measuresyou can save over two-thirds ofyour water consumption. In this case,however, we have an eaves solution that is dead cheap and looks verynatural. After all, one ofmy personal mottos is semper ut funquet, let italwaysbefunky. (The otherone is Why messabout?)We have, therefore, a system of eaves that has as its main aim thereta ining of turves on the roof, and protecting the edges of the outerturves from drying out too quickly. lt also has to prevent water reachingthe outer walls of the house, and to prevent water from soaking the

straw insulation layer in the roof. Even without a gutter system, it wassti ll quite fiddly to do, and took several days.The first task wa s to extend, where necessary, and make good the

41


49/121

canvas layer on top of the willow laterals, and to contain the straw layeradequately. I used several ex-army blankets to achieve this. In someplaces there are small gaps where a small bird or bat could get intothe straw under the eaves and make a nice cosy nest there. That 's OKby us. We humans have deprived so many of our fellow creatures ofthei r natural habitats that the least we can do is to put new habitatopportunities into every thing we build. You will see from diagram 9that this underlayermeets the rubber at theouteredgeoftheeaves.The two laye rs are folded together and pinned at about 2ocm intervalsto the edging logs, which are linked loosely to each other and to therafters by means of short lengths ofgalvanised wi re and staples. Thismight seem a very rough and ready system to you, dear reader, but weare ta lking very funky 3D here, and it works. The factor that I have sofar not mentioned in this design is the slope of the turf roof at theedge. We have placed straw bales up to 1m from the edge. These balesare one foot (3ocm) thick, and the roofis sloping. The challenge is tofill the triangular space ('x' on diagram 9, page 34) with straw to give

42


50/121

the most gentle slope possible so that the turfwill stay on. In practiceI found the best way of doing this was to combine all the folding,stapling, cutting and straw-stuffing in short stages of about twometres at a time, working round the perimeter. I measured the nextlength where one log cou ld comfortably touch a number of rafters,from the first to the last rafter. This might be anything from 2 metresdown to a tr icky bit where a short length ofonly a foot {Jocm) wouldbeneeded . l cuta logtosizeand offered ittotherafterstocheckthetit. I then cut four or five handspan lengths of wire, and stapled thewire one piece to each end of the log, and two or three to parts of thelog that came nearest to the rafters underneath. Then I stapled thewires to the rafters, and one wire to the last edging log. Then Igathered together the canvas with the rubber membrane, folded themtogether, and pinned them at intervals to the log, just over the top ofthe log ifpossible to delay the action ofwater making its way throughthe na il hole to rot the log. Am I boring you? OK let us proceed. Iwasthen able to stuff the space 'x' thus created wi th new straw takenfrom the 30 un used bales, and th is procedure, bit by bit round thehouse, used up 20 of them

One thing Idislike about modern buildings is all the straight edgesthat are so unnatural. I wanted an edge to the roof that had textureand wavy edges, and I am very happy with the look of the finishedresult. I reckon that the edging logs used here {mostly offcuts fromsecondary rafters, approx 4"j1oomm diameter) will be the first thingsto rot in this house, but so far none have. When I am out cuttingfirewood and I find a Jog covered in moss, I bring the moss back andstuff it on or around one of these edging logs. So maybe, if the mosstakes and builds up a good layer, Iwon't need to replace the logs whenthey go, a lthough that would be easy to do. Rob Roy, the Americancordwood masonry builder and writer. started edging turf roofs withhuge ex.railway sleepers. Hi s latest buildings are edged with clumpsof moss. Much easier on the back. I am also hoping that the variety offru iting and other plants in the roof will in time build up a growingedge that transforms the drudge ofbuilding maintenance into the muchmore pleasant gardening tasks ofweeding and pruning. The roofwasnow readyforturfin'.

43


51/121

T u r ( i n ~l'msorry,llovedtheBeachBoys.lcan'thelpit;everytimelslarttalkingabout tu rfing I start singing surfin' songs and making awful puns.Compassion, dear reader; that's alii ask.

Whyturf, rather than, say, sh ingles or ti les or slates?Well, for one thing,I honestly believe that if our ancestors had invented a way of makingsingle large waterproofsheets that were oforganic materials and lasted30 years or more, the whole look of our civilisation would have beendifferent. Most modular roof systems, like slates, were a very timeconsuming, heavy and expensive way of trying to keep wate r out over areasonable surface area, and ~ e d a steep, flat slope with strong inflexiblesupports to hold them up. Hence our ideas of vistas of pointed roofs,gable ends, and a ll the paraphernalia that goes with what we think a groupof'houses'looks like. We are almost stuck with this shape. based thoughit is on technological limitations of he middle ages. (Plannerswill usual lyonly give permission for a 'house' with allthe culturaland obsolete baggagethat the idea of a house entails. Nowadays, ofcourse, a 'house" usuallyhas a 'garage' next to it, or built into it. Will planners of the 23rd centurystill be requiring country 'houses' to have a traditional garage attached tothem, even though cars are long obsolete?) The most organic system ofour ancestors for keeping the rain out was thatch, which looks good,provides habitat, is renewable, and is warm if you can get it in sufficientquantities, but only lasts a decade or so before it needs replacing. Imagineif rubber pond liners had been available 200 years ago. Our expectationofwhat a human settlement looks like might be more like a badger settwith skylights. Technology doesn't have to be big and hard and shiny andfighting nature. Technology could lead us back to a more harmoniousrelationship with nature, and if this house demonstrates only that, itsconstructionwill havebeenworthwhile.

Formetheessence ofa turfroofis thefactthat l haven 't taken anysurface of the earth away from nature I've just ra ised it up a bit.Plantsandcreaturescanstilllivethere.ln fact, sinceit isn ' t grazed orprone to bracken invasion, it provides a new, protected habitat. Theother reason to go for turf is for its low visual impact, which for me isnot an inte llectual concept or a planning buzzword - it's just natural.How many other anima ls , except humans, can you think of that build

44


52/121

theirdensonthetopsofhillsandinotherplacesthatstickout for miles and announcetheir presence to the wholeregion? The more I live in anaturalsedudeddenthemoreI see these loud, rectangular,mock-permanent humanliving zones as a seve reaberration. More about thislate r, maybe, but for now letusjustsay thatfor meithasto be turf. With a few tay-berriesand strawberriesinit,of course.

There are some goodbooks and leaflets on turf roofs (see appendix). Some people preferto chuck loads of earth on and then seed it. We prefer to cut ind iv idualturves from nearby, load up a wheelbarrow, hand or carry them up tothe roof and lay them then and there. In the long run it doesn't makemuch difference . The celebrated architect and turf roof maker,Christopher Day, who lives just over the hill, once told me that theturf roof on the Steiner Kindergarten school at Llanycefn, made inthe early 1980's, has been found to have 23 species ofgrass, none ofwhi ch was there in the first place! So a turf roof is a living thingwhich evolves to suit the conditions in which it finds itself.There are a few factors to consider in turfin'. Slope, holes, roots andwater. The rest is just hard work with a good view.When grass roots grow down and hit a waterproof membrane, theygrow sideways and mesh together in a strong web. At first, however, forthe first few weeks, each turf is on its own. A new turfwill slip off a smoothsurface if a lot ofrain saturates it and the root system is still inadequate tobind the whole section together. So getting a healthy turf to stay on is aquestion of grip and root system, rather than a simple slope formulaSome texts will tell you that a turf roof must not exceed a 10 % , or 15%,slope. Although youwill have fewer problems with a gentle slope like this.

45


53/121

there are cliffedges on the Pembrokeshire coast a cOuple of miles fromhere where grasses and wild flowers are clinging on at angles of over45%, indefianceof all thetextbooksFor a roof shape, I personally prefer the 'mushroom' kind oflook wheremost of the roof has very thick insulation that comes down to nothing atthe eaves,therebygivingasharperinclineattheedges. Theway)aneand Ita ught ourselves to apply turves allows us to lay them on quite a steepslope at the edges. We start at the edge, and work towards the centre. Theedging turves are twice as long as the more central turves. For a normalturf, we havefound about a square foot (3ocm x3ocm) area, and about fourfingers depth of soil (like the Miwok) is about right. If you are bu ildingsomethingwi th seriously strong roofmembers, say oak, or pine beams 10"x 3" {25 x75cm), then you can use considerably thicker turves. This will giveyou slightly improved insulation (but not much soil is not brilliant asinsulation), and greater potential for d rought resistance and deeper rootingplants. We don't send back a turfif it is 6" deep (t&m), but a seriesof hemplays heavily on your arm and backmuscle5. An edging turf s twice as longbut the samewidth and thickness, and we use these for the first two rows atthe edge. One person can just manage to carry one of these, and you willget about six in a wheelba rrow. For cutting. use a sharp spade (yes, you cansharpen a spade on a grindstone. lt makes a huge difference. Mind yourtoes). When you have a large number of turves to cut, i.e. approx. 1400


54/121

turves on this roof, it saves a lot oftimeto make a standard size and to befairly methodical in the cutting. Jigsaw puzzles are fun, but not on this scale.Soweplanout thearea, andwtin a sequence.

Before laying the turf, it is a good idea to place a newspaper or someabsorbent fibrous material such as o ld blankets on the rubber. You canbuy 'geotelrtiles' for this purpose. This serves the purposes of protectingthe rubber from small pointed stones or thorns that might pierce it whena person walks above, and of providing a medium for roots to form a matin. On some parts of his roofwe did not put this newspaper lining (severalpages thick), and at times the next summer the turfbecameverydry. So Irecommend that you do it all over.

just a note on holes. Make sure not to a llow a single hole to go unrepairedbefore turfing. One day while Iwas turfing a whole lot of nails spewed frommy pocket onto the rubber and Itrod two into it before I noticed. 18 monthslater we noticed two damp patches in the canvas ceiling, and needed tospend a morning chasing up and puncture-repairing those two holes!

Laying the turves is excellent exercise. We started wherewe could reach,to get into practice. The edging turves are laid lengthwise up the s lope, sothattheirrootstructureholdsthemon thesteepestbit.lmakesurethat thebottom end of the turfsits firmly in the depression behind the edging log,and thump it into place with my fist. As we moved round the eaves to thesouth part which is about z.smetres above the ground, it was a matter ofclimbing a step ladder with a long edging turf, cradled in such a way that (would fall into place with it, aiming at the exact spot. Any turf which splitdown the join was used as two norma l turves for the next or third row. Afterthe first two rows were complete the roof started to look, especially bymoonlight, as I had imagined it. I can remember dearly the pure thrill ofseeing this organic dream taking shape and becoming real. For the rest ofthe turves itwas a matterof patience, arm muscles and good helpers. Mostpeople in the community helped with a few hours ofturfin'. Here' s Marthahelping )a ne. For the normal turves. we devised a slingof rubber thatwouldreach down to the ground and which held two or three turves. An efficientway ofoperating is to have one person digging, one person transportingand another laying theturves on the roof. By the end ofJanuary 1998 we gotto the top circle, where the rubber was still uncut, stretched over the roofhole like a trampoline. Turf's Up. On to the skylight.


55/121

The Skyl ightThe skylight does not winch up, to let out smoke from a central fire,despite several people's romantic suggestions, although we did try itfor a couple ofdays. Too much smoke stayed in .As a vent in hot climatesthis would be quite feasible, though. In Wales, however, keeping coolhas proved to be fairly easy. The main function of the skylight is as afabulous source ofdaylight, beaming down into the centre of the housein all weathers. Iam typing by its light now, with a terrible wet and windyjanuarydayoutside,andwith no artificial light ing.In composition the skylight is two slightly conveK coach windows,each measuring4ox 8o inches (approx 100 x2oocms) and 6mm thick,laid one on top of the other with a rubber seal2mm thick, sealed withs ilicon, between them. it's the best value double glazed skylight I canimagine. (The rubber. ofcourse, came from offcuts.) The bottom windowis laid over the hole, which measures 34" (86cm) in diameter, straightonto the grass. Turves are piled under the ends and on top of the topsheet to hold it down and to eliminate any tendency for the wind to getunder the glass and make offwith it .

48


56/121

(diagram 11) Cross sectian afskylight.Underneath, offcuts of canvas are tucked over and under the e ~ p o s e dstraw bales to tidy the hole up. At the top of the hole, where the rubbermembrane is close to the underside of the glass, I wrapped a piece ofspare tubing around the straw circle a few limes, supported on woodenpegs driven into the straw, to raise the rubber 1" (25mm) or so to touchthe glass, then sealed the join all round with silicon sealant (diagram n)Actually, there is still a slight air gap of a few mm on one side, which wedon't mind as it allows for a bit of fresh air to make its way in up there.When there is a gale the rubber loosely touching the glass occasionallytreats us to a sound best described as a giant raspberry.The on ly thing to add to this is that we were especially care ful to seethat the sheets ofglass were immaculately clean on both sides beforewe laid them on the roofand sealed them together, and that we did notget any sealant on the bit we wanted to be completely dear. Easier saidthan done- coach windows are heavy things. just a tip on getting holdofbig windscreens or toughened glass windows like this, ifyou are likelyto want to: it's easier than you wou ld think. I got seven of these; four forfs each and three for free. On a new coach, each window is toughened,convex. thick glass of high optical quality. They cost over f2so each.When the coach is eventually scrapped, the windows will probably be inas good condition as when they were new, but obsolete. Unless some

49


57/121

crazy permaculturalistwith nothing better to do than build funky housesand greenhouses comes along, they are virtually useless. But few depotmanagers are willing to chuck out such beaut iful artifacts as these thatcost so much and must, surely, have some value. Sol have, on severaloccasions, located algae-covered s tashes of lovely wi ndows andwindscreens in bramble patches behind junk yards and bus depots afterdetailed cross-examination of the people working there. Don't ask theboss first; ask the guy ma intaining the vehicles, then offer the bossmoney. No more than a five r each. though - we don't want inflation toset in to the used windscreen marke t, do we?

50


58/121

WallsThe walls of his roundhouse are mainly made ofwhat is called cordwoodmasonry. lt is, apparently, the first house in Britain constructed in thismanner. The claim for first buildingofcordwood goes, as far as I know, toBen Law, the Sussex permaculture designer, for his cordwood and turfworkshop. (Ben and I have something else in common, which is that weboth have almost iden tical African sty le wooden chairs, made by each ofus on a permaculture camp from two halves ofone beautiful oak plank)

There are no inner walls, so this chapter refers to the outer walls. Forthe sake of clarity, I will call 'a wall' that section of wa ll hetween twouprights, although around 6j13 of he perimeter, from the West throughthe North to the East, it is really one stretch ofcurved wall, and most oftherestiswindow.

I started without the benefit of having read more than about twopages of text on cordwood masonry. In Ken Kern's book The OwnerBuilt Home there is a fuzzy photo of an old guy in the Mid-West besidea wall made of log ends that he was covering with chicken wi re beforerendering. I have been wanting to try out this method for over twentyyears, since seeing this photo. The Short Log ond Timber Building Bookby )ames Mitchell gave one useful clue: to put a layer of insulation inbetweentheouterand inner mortar layers. Everything else I made up,but judging by a read of Rob Roy"s book Complete Book ofCordwoodMasonryHousebuilding (which I have subsequently obtained, in exchangefor a wooden bowl, from the author when he paid us a most e njoyablevisit), I hit upon reasonab ly acceptable ways of doing it. The maindifference from the straight American system is that we had resolved touse no cement in the construction of this house, so in place of Rob'smortar ofcement, sand and damp sawdust, we used raw mud from twopaces outside the eaves the big pi le of subsoil that came out of thebank that this den is set into. The earth was high in clay content, withsome sand and grit; in fact, as far as I can see it merits the term 'cob".We mixed it by the barrow load with water from buckets placed underthe eaves (it rained almost every day) and three or four handfulls ofsh redded straw or dead bracken. The insulation is straw. Robrecommends sawdust. which would have done equallywell, maybe better,

51


59/121

but straw works fine. The log-ends are all as near as Icould get to 16"j4ocm without measuring. Icut maybe two hundred at a time with theelectric chainsaw up by my workshop and brought them down by thetractor and tra iler. lt is good to have a big pile of logs to choose from,so that you can keep ringing the changes in si.ze and texture. Thereshould always be the perfect log-end waiting to be laid in the nextspace. I it's not there, you need to cut another load.I s ta rted with the wall in the North-West; the one next to the frontdoorjwindow. This would mean we could work our way clockwise(sun-wise for New Age readers) all round the perim

building a low impact roundhouse

Documents