1 IIMS NEWS

Being the newsletter of The Institute of Information & Mathematical Sciences, Massey University at Albany Edited by Andrea Babylon

Inside: My Life in a Dustbin pg 2 Loud Shirt Day pg 3 Computer Science News pg 4 Kebnekaise pg 5 From t˙ Head of IIMS Tahuna Conference pg 10 From the Head of IIMS Acronyms and Names pg 11 Telegram-hacking Scandal pg 13 Dear all, Te Kooti’s Pa pg 14 Snippets of History pg 17 Well this truly is the final IIMS newsletter, but like all Wood Pigeon pg 21 of the good things that IIMS does you can expect it to continue in the future in a substantially familiar format, probably on a tri-annual basis.

I’ been a member of IIMS since November 2004 (even while within NZIAS) and there have been a lot of changes particularly among the staff. Shaun and Carlo, relatively junior staff then, are now Associate Professors. The group in information systems has gone. There have been other promotions and appointments - the arrival of Marti and her group are of special note. Manifest suc- cesses too - Mick’s FRSNZ, Robert’s ANZIAM medal, other Fellowships and a good number of Marsden grants (10 or more I guess). Lots of other successes, grants, prizes and awards for IIMS staff over those years - go back and read the newsletters! IIMS has been success- ful, and now it will part of something which I believe will be even stronger and offer those of us here at Al- bany an even greater opportunity to determine our own future. Just for the record (or if you happen to have been on long leave to Mars .) IIMS, apart from the staff in IT, will be joining with INS to become INMS formally from 1 January, but in effect from 1 October

Continued on page 2 2 this year. The IT staff will be supported by we can do as the “Sciences at Albany”. As INMS until exactly where they finally set- I have repeatedly said, our strength and op- tle within SEAT has been decided. We had portunity lies in a multi-disciplinary focus, a wake and people spoke about the past, it unique in New Zealand. We have strength in was good to see many old (?) faces. But the all the sciences, but little by way of “critical future beckons. mass” so no discipline can go it alone. I’m sure this strength will be recognized in the The IIMS Postgraduate conference ran along PBRF round to be announced in April. We very smoothly this year. This is a really great will lead Massey and the country in a number event on the calendar of IIMS. Much thanks of areas. are due to Freda and Amjad for the running of this event (which will only be bigger next So my final words as IIMS HoI are an exhor- year). These things don’t run themselves! Of tation to get involved with making the new course Freda will continue in her role sup- Institute the best it can be. porting the Postgraduate activity in the new Institute. I guess we were both pleasantly Cheers Gaven surprised to find out recently that our com- bined Institutes had over 75 PhD students Gaven Martin (!) so there will be a lot more variety next year (and bigger prizes ! On that front did anyone see Shaun eat the unclaimed prizes he tried to give out ?). I didn’t get to go to many of the talks, but I did read the posters My Life in a Du

Loud Shirt Day

Loud Shirt Day

Figure 1: From left to right: Mr Yan Ou, Dr Winston Sweatman, Dr Cathy Hassell-Sweatman, Freda Mickisch, Dr Barry McDonald. Photo sourced by Annette Warbrooke. 4

be playing a key role as we refine Computer Sciences place in the teaching and research nexus. Andre and Napoleon Reyes have also made great headway this year in developing Computer Science News a new software infrastructure for the robot soccer system. This has proved a valuable Computer Science News platform for student projects and this soft- ware gives CS’s Computer Vision group a News from the Computer Science Group new area for research growth.

Ken Hawick has further enhanced his al- Congratulations to Elena Calude who has ready formidable international reputation by been promoted to Senior Lecturer. Elena adding a British Computer Society Fellow- has recently published an important article ship to his Fellowships of the Institute of on “The complexity of Riemann’s Hypothe- Physics and of the Royal Meteorological So- sis” in the Journal for Multiple-Valued Logic ciety. Ken also achieved another milestone and Soft Computing 18(3-4) (2012), 257-265. as he recently completed his 200th refereed publication. In addition, he has taken on Daniel Playne has taken over the director- two new PhD students Tim McMullen (co- ship of the Centre for Parallel Computing supervised with Chris Scogings) and Jeremy Research (CPC). As many will know al- Willemse (co-supervised with Martin John- ready, CPC has been repositioning itself to son). Tim and Jeremy are both working on focus on research into parallel computing al- advanced graphical simulation systems. gorithms, parallel languages and especially data-parallel computing based on the use of In May, Chris Scogings took over from graphical processing units (GPUs). Martin Howard Edwards as the Programme Direc- Johnson had led CPC to successes in grid tor for the Bachelor of Information Sciences. computing but Martin is now focussing on The “new look” BInfSc now offers majors in energy-efficient parallel computing. Martin Computer Science, Information Technology and Ken published a peer-reviewed article on and Software Engineering. Chris ceased act- this earlier in the year. ing as Institute Timetable Officer only to find that the new duties of Programme Directors In 2011, Napoleon Reyes went to Eastern Eu- included checking the timetable! Despite the rope, for his first sabbatical. His initial plan rigours of timetable checking, Chris managed to spend the sabbatical in Toyota, Japan to fit in the successful supervision of Delan was hindered by a staggering combination Ren who has completed his MSc thesis enti- of earthquakes and tsunami, followed by a tled “An Ant Colony Simulator”. Chris has nuclear blast. It only goes to show that also blended his research area of agent-based we can always make the best of efforts to modelling together with his love of military have a grand “plan A”, but in the end may history and produced a paper entitled “An still have to resort to “plan B”. Nonethe- Agent-Based Model of the Battle of Isandl- less, his sabbatical gave him lots of unex- wana” to be presented at the Winter Simu- pected opportunities: he delivered his first lation Conference (Berlin, Germany) in De- invited keynote speech in an international cember 2012. workshop in Hungary, had his first TV in- terview appearance in a documentary film, Andre Barczak has taken over Chris’s role as entitled “Dnes v buducnosti” (translated as Major Leader of Computer Science and will “Today, in the future”) in Slovakia, an oral 5 presentation in an international conference Mitchell is studying for his PhD in GPU in Spain, and he also wrote a book chapter simulations with Ken and Chris, and Vic- publication regarding his research on path- tor is completing his Honours year and is planning using hybrid intelligent systems. working on a simulation project with Daniel Playne. Both Mitchell and Victor are veter- Ian Bond and his PhD student Michael Zhao ans of IIMS’s summer student project pro- are continuing to make great headway ex- grammes. ploring planets as part of the MOA astro- physical team. As well as contributing to Two other IIMS summer scholar veterans - Massey’s increasingly successful first year pa- Alwyn Husselman and Tim Lyes have also per on astronomy, Ian and his team are also had successful years, achieving several peer- making continued use of parallel computing reviewed publications each. Alwyn is work- facilities and many of our students will have ing on agent-based modelling with GPUs been taught paralllel programming by Ian. and Tim is working on stereo and advanced graphical visualisation. Both are working Teo Susnjak completed his PhD thesis enti- with Ken and Chris. tled “Efficient Boosted Ensemble-Based Ma- chine Learning in the Context of Cascaded Looking forward, Computer Science will have Frameworks”. His thesis has been placed on two IIMS-funded summer internships this the Dean’s list - reserved only for theses of ex- year, that will support students developing ceptionally high quality. Teo completed his simulations on mobile tablet computers. In- thesis just in time to take up a lectureship in terestingly, GPUs are now becoming avail- Information Technology at the Albany cam- able in mobile tablet devices and needless to pus. Congratulations to Teo and his super- say these need to use energy efficient pro- visors, Andre Barczak and Ken Hawick. gramming techniques to prolong battery life.

In 2012 we introduced the new first year elec- Ken gave a recent IIMS seminar on Domain- tive programming paper 159.171 which al- Specific Programming Languages, and we lows students to get a taste of programming are also hoping to have an additional sum- in the python programming language. De- mer student team work on developing some spite introduction of this new programming domain-specific programming languages and paper we have seen record growth in CS stu- apparatus for GPUs and tablet computing dent numbers for our mainstream C/C++ systems. first year programming papers 159.101 and 159.102. Special thanks to Ursula Scogings The Computer Science team are planning and Peter Kay and to all the CS postgrad further developments in GPU computing and student tutoring assistants who contributed Ken and Andy Gilman are developing a pi- to the often times hectic first-year teaching lot study into the combined use of Field programme. It is a testament to these pa- Programmable Gate Arrays (FPGA) with pers that our CS students seem to find no GPUs. In 2012 Andy also consolidated the difficulties picking up employment. successful new CS Honours paper on FPGA programming 159.703. This is expected to In July Computer Science students Mitchell grow even further in 2013. Johnson and Victor Du Preez presented no less than two peer-reviewed confer- As many of you will know CS will be join- ence papers apiece at the WorldComp ing the Sciences@Albany endeavours in Prof multi-Computing Conference in Las Vegas. Gaven Martin’s Institute. In addition to 6 our present scientific collaborations in math- and two half days in the Swedish Arctic (Sun- ematics and statistics, there are other ex- day 15 July - Friday 20 July, 2012). citing areas of complex systems and compu- tational biology, computational physics and Last European Summer, I had a few days computational chemistry that we all look for- break in Sweden and managed another visit wards to collaborating in. No doubt 2013 will to a place on my list of places to go to from be even more interesting than 2012. my teenage reading. I travelled to the north of the Arctic circle to tramp near and climb Chris Scogings Kebnekaise (Giebmeg´aisi)the highest moun- tain in Sweden.

I travelled north by sleeper train from Stock- holm (a journey of 16-20 hours). I was keen to see to the end of the line and so went all Kebnekaise the way on the train to Narvik for a couple of hours in Norway, before catching the train Kebnekaise back into Sweden to Abisko, ready to start walking at 5pm. The Kungsleden and Kebnekaise, four full

Figure 2: View along the Kungsleden 7

Figure 3: View along the Kungsleden

After lumbering along with my pack for tops of the hills. As in Britain, the Swedes about 3 hours (14 km), I reached Abiskojaure commented on it having been a very wet hut where I spent the first night. year.

My route was along the northern stretch My first full day I walked to Tj¨aktjahut (33 (heading south) of the Kungsleden which is a km). The route follows up a long valley. I multi-day Summer walking route and Spring was lucky with my timing as flooding had ski trail. This has comfortable huts (that I closed sections of the path the day before. used) or the option of camping. Gas stoves The weather now was pleasant (neither wet are provided, and I lived primarily on cous- nor cold). cous with tomato paste in the evenings and muesli in the morning. The scenery is a little The following day, as I continued over a pass, like Scotland (I ascended Meall nan Tarma- I came to S¨alka hut where the warden said chan just over a week before) but the hills that the weather did not sound overly hope- and valleys are on a larger scale and there ful for the Thursday when I had hoped to was still reasonable amounts of snow on the climb Kebnekaise. I might have taken a 8

Figure 4: Ridge returning to the main summit of Kebnekaise from the North Top at about 1am (with a few snowflakes). short-cut back route direct to the base of the I discussed my plans. The mountain station mountain here, but decided that it was bet- provides a guiding service to the summit and ter to continue on the Kungsleden as planned so they are aware of conditions on the peak. and I completed the 24 km to Singi hut. The large amounts of snow made crampons unnecessary. Most people were just using ski At Singi hut a side path leaves the Kungle- poles but I felt more comfortable having an den to the mountain station (large mountain ice-axe with me. The weather on the radar hut) at the foot of Kebnekaise. This 14 km for the morning was still not getting any bet- took most of the morning. When I arrived ter and so I made use of the arctic summer I checked in for a couple of nights and con- light and set off at 5pm to climb the moun- firmed that wet weather was still anticipated tain overnight. the next day. I had a brief lie-down while I waited for the hire shop to open at 4pm. As I ascended, during the next four hours, I met streams of people coming back down The man at the hire shop was most helpful as from the top. A rain shower also passed and 9 waterproofs were donned. Apart from the next day, it was the only rain I remember in Sweden. As I went up it got colder and colder. I crossed large patches of snow for the first time in many years. A short way below the summit, after some searching, I found a semi-abandoned half-snow-filled hut. Wet clothes were removed and replaced with warm stuff and I consumed some food. It was cold and I was glad to be off on the move again.

Up at this higher level, late evening, local vis- ibility was fair but it was in cloud. The twi- light was reasonably light, for although too late for midnight sun, if I could have seen it, the sun would not have set for more than about half an hour.

I trudged up a runnel of compressed snow to the summit, passing a group of three coming down at about 11:30pm. The summit was white, quiet and lonely. A likely contrast Figure 5: Below the summit of Kebnekaise, with daytime, I had it to myself. As there just before midnight was a large deposit of soft snow, I was able to follow footsteps left by a guide in the ridge By this stage my knees were reasonably tired crest to the northern top. This route would and I was slow on the final stretches into the usually be a much harder icy alpine chal- valley. I did not arrive at the mountain sta- lenge. The north top is more solidly rock and tion until about 7.30am, a bit over 14 hours will overtake the main summit (about 2104m since I had left. I was just at the right time depending on snow) if too much global warm- for the morning opening hours of the hire ing occurs. shop to return my ice-axe.

In the midnight light, I returned past the The Thursday as predicted was very wet. summit to start descending a little after 1am. During the day, several people got wetter and I went to the other high refuge on my way saw less of, and on Kebnekaise than I had. I down but did not go in as someone was sleep- sought food, a shower and my bed! I also ing there! After some more food sitting in the enjoyed the Sauna later in the day. cold I descended comfortably down the snow slope seeing the sun and some views at about My final morning was the 19km hike to 3am. I was somewhat too far to my left and Nikkaluokta and the lunchtime bus to Kiruna so traversed back to the right to meet the and the railway. It was a steady plod and I main path marked by red paint splashes. could appreciate why so many folk who had been up the mountain celebrated by taking The route back was the same as coming, it either the helicopter all the way or else a boat passes over another peak that was a hard part of the way. However, at least I had just climb back up but even harder coming down. had a rest day and at this stage was starting 10 to get fitter from the earlier tramping in the next morning and we were introduced to trip. our overseas speakers who were working at the forefront of their field. That evening Winston Sweatman we had a special reception celebrating the work of Professor Peter Whittle, a famous New Zealand statistician, with three more speakers. After seven talks my brain was overloaded but I slept very well. The week went quickly with four talks in the morning and a late lunch. In the afternoons we went Tahuna Conference for walks down the beautiful Tahuna beach, caught up on some maths, rested or looked Tahuna Conference for free wifi (yay for McDonald’s). The mo- tor camp is the biggest in the southern hemi- What I did in my summer holidays. sphere with 50 acres and 1500 campers so they organized activities in the big tent in the On a wet blustery Auckland day in early evenings. We had a maths group in the trivia January we caught a plane to sunny Nel- quiz night and spectators but no participants son. We were off to the NZMRI summer at the karaoke night. Wednesday was our ex- workshop, the topic “Random walks on ran- cursion day. A tramp was organized to the dom graphs”. I had to look up the defini- top of a local mountain and although they tions on Google so didn’t know what I had all enjoyed it there were some sore legs and let myself in for. Anyway, after deposit- blisters the next day, others went into the ing our bags at the Tahuna Beach Holiday Able Tasman National park for a walk and Park where the conference was to be held we we headed on a ferry to beautiful Kaiteriteri headed in to Nelson city. It is a lovely city with its golden sands beach. We had a per- with real streets. We discovered the Satur- fect day without a cloud in the sky. That day morning markets with all the different night we had a barbeque. There was a good stalls selling various foods to arts and crafts. crowd from Massey Albany with Prof. Gaven We sampled some of the many berries on Martin, A/Prof. Shaun Cooper and A/Prof. offer. We then found a second hand book Carlo Laing representing the staff and Josh, shop, walked in the beautiful Queens gar- Irja, Graeme and Lynette, the students. We dens, walked through the art gallery and all had a thoroughly enjoyable time listen- loaded up with some groceries from the su- ing to talks from our overseas visitors and permarket before unsuccessfully looking for exchanging ideas. I have to admit to not un- a bus to get home. A local told us “Nel- derstanding a lot of what was going on but son didn’t have buses or bus stops and that it was good to see the tools we have learnt it was only a short walk along The Rocks in various undergraduate papers being pulled road home”. (I don’t think he had ever actu- together and used in new ways and good to ally walked it!!!) With loaded back packs we meet new and interesting people and the pre- walked and walked. It was a good hour but dicted bad weather for Nelson didn’t happen an enjoyable walk with nice views. We un- so it really was like a good summer holiday packed in our cabin and met other students. with a lot of intellectual stimulus. Mark Holmes from Auckland University, our organizer, did a catch up tutorial for us that Lynette OBrien 13/1/12 evening. The course started in earnest the 11

Figure 6: Josh Duley, Shaun Cooper, Raazesh Sainudiin, Irja Geraets, Matthew Parry

the moment.

Many of you will have heard discussion on GPUs - a GPU is a graphical processing Acronyms and Names unit. Why is it called that? Well a GPU is not really a stand-alone processor that can Acronyms and Names do useful work by itself. It is a specialist ac- celerator, sometimes called a co-processor A colleague recently commented that one of that assists a CPU to do particular opera- the problems in understanding a seminar in tions. GPUs started off just doing graphics another group’s area was that there were so related calculations, but research groups, in- many acronyms. It is true. We end up work- cluding our own at Massey, found ways to use ing inside our own fortresses of specialist jar- GPUs to do other data parallel calculations gon, names and acronyms that we tend to and to speed up programs that sometimes assume colleagues are familiar with. I there- had nothing to do with graphics. fore thought it might be interesting to reflect on some of the acronyms and name tags of But what is a CPU I hear you cry. Well 12

CPU stands for central processing unit. Why What field? Well in this case “in the field” central? Well in the bad old days (before just means by the likes of us - in other words my own time I have to say) a computer by end-users and not by the manufacturer in was made up of separate processing elements, the factory. FPGAs are usually very good at and in some cases the processor was made data-parallel applications of a certain type up of individual discrete transistors on sepa- and so are likely to be used as accelerators rate circuit boards. There were (non central) or co-processors. processing units that took care of peripher- als like keyboards and the like too, and the Our group’s interest is in trying to make whole assembly was quite likely spread over CPUs, GPUs and FPGAs all cooperate to- several boards or cabinets. Not like the sin- gether to make certain applications - simu- gle CPU microprocessor chips we are accus- lations or vision and image processing appli- tomed to today. cations for example, run really fast. Faster of course means we can do things that other To complicate matters however, our single groups cannot. Sometimes having faster de- chips now often have multiple processing vices means that a bigger problem is feasible cores on them. A core just means a pro- or that the parameter space of a model can be cessing element that can work mostly inde- explored interactively rather than overnight. pendently (in the case of CPU cores) or not In Computer Science, faster always means very independently at all (in the case of GPU better. cores). The number of cores available on CPUs and GPUs is increasing. At the time “Computer Science” itself is an interest- of writing it is very common to have 2, 4, ing name-tag. What does it mean? Well we 6, or 8 fairly powerful cores on a modern don’t necessarily get to define it, as it has CPU like the Intel Xeon CPU. You may hear a fairly well known international meaning al- some of us getting excited at the prospect ready. Nevertheless we do get to put our own of CPUs that have 16 or 32 cores on them. spin on it. I always think of it as meaning In Computer Science at Massey we are al- what it says - the science of computers and ready experimenting with AMD Bulldozer computing. A subtle difference arises when processors. AMD like many vendors are good we consider the name-tag “Computational at coming up with strange but memorable Science” - which most unfortunately has the names for their devices. As an aside, one of same acronym of CS. Computational Science the most amusing namespaces in computing is usually taken to mean doing science of all at present is the one that Google and co give sorts with computers and computing appara- to different versions of the Android operating tus. It thus takes in computational physics, system software. They use names like “Ice computational chemistry, computational bi- Cream Sandwich” and “Jellybean” and ology and so forth too. the forthcoming “Key Lime Pie” - suppos- edly dessert-oriented names! Just to add to the confusion, at Massey our Computer Science group also works on Another of our current favourite acronyms is computational science problems and applica- FPGA. Heard of that? FPGA stands for tions. Our particular speciality is in Complex field programmable gate array, and they are Systems simulations. Oh dear - yet another also special purpose processing devices that over loaded CS acronym. Never mind, I am (usually) consist of an array of switching - sure it all becomes clear from the context just ements or gates, that can be programmed or which “CS” is which. I am also sure every- configured at the hardware level in the field. one knows what science means, and that no 13 one confuses science with technology or en- UK, involving the news media, and, possibly, gineering. political corruption, is not really new! Sim- ilar events took place in NZ well over 100 Complex Systems come from all sorts of years ago: albeit the technology of the time contexts, including the above-mentioned sci- was slightly different. The following case re- ence areas, but also from networks, the In- ported in Appendices to Journals of the NZ ternet itself, infrastructural networks such as House of Representatives (Parliament, in to- power grids, and even from financial, eco- days terms) relates to accusations that sen- nomic and sociological systems. We usu- sitive telegraph messages were being inter- ally describe these as “complex” to reflect cepted by telegraph operators, passed on to that there is some emergent or “unexpected” unauthorised parties and then used for polit- complexity that comes from the interactions ical purposes. within the whole system and not just from the microscopic constituent parts which may “SEVENTH ANNUAL REPORT OF THE individually be quite simple. TELEGRAPH DEPARTMENT, NEW ZEALAND [to the NZ House of Representa- Some of these names and acronyms have tives, dated 5th August 1871] been around a while. Some will come and go and be forgotten. I am reasonably sure CONCLUSION. GPU has a definite shelf life. It is useful this decade, but like other name tags that some- The Report for the present year can hardly times seem terribly important to us at the be complete without a short reference to the time will perhaps be forgotten by 2020. The “Telegram Libel Case” which lately attracted ideas of GPU and FPGA computing will pos- some notice. A newspaper in Otago [Otago sibly just become subsumed by future gener- Daily Times] having published certain arti- ation CPUs. While “CPU” has some histori- cles charging the Telegraph Department with cal baggage it is so well established now that corrupt practices, the Government instituted I think the term CPU will still be around for proceedings against the reputed Editor. Af- the next 50 years. Possibly the optical or ter a long and minute investigation before the quantum technology that they are based on Resident Magistrate at Dunedin, Mr. Barton by then may be unrecognisable to us in the was committed for trial on the charge of pub- now, but the name and concept will be the lishing a false and malicious libel. The ob- same and will live on. I am sure the same ject of the proceedings - namely, the disproof will be true of Computer Science itself! of the charge made - having been thereby gained, the Government desired to spare him Ken Hawick from being brought to trial. It seems that he was afterwards advised to take proceed- ings against some members of the Govern- ment for their original prosecution, so that it is necessary to refrain, for the present, from further comment on the case here; but Telegram-¯¨ing Scan˚l it may be added that he also brought a charge against Mr. Lemon, the General Man- The Great NZ Telegram-hacking Scan- ager [of the Telegraph Department], of hav- dal (1871) ing improperly divulged a telegram to Mr. Gisborne, the Minister then acting as Tele- The current phone-hacking scandal in the graph Commissioner, which charge the Mag- 14 istrates summarily dismissed without even ten quoted model. They find Ministers not calling upon Mr. Lemon’s Counsel to reply.” guilty, and recommend them never to do it again. It will be observed what emphatic (Note: Charles Lemon was an early General language the Committee employ when Gov- Manager of the NZ Telegraph Department ernment are in the right, and how extremely and served in the role for many years. The moderate and soft-spoken they become when distinctive wire-draped poles that appeared Government are in the wrong. They evi- around the country during his tenure were dently entered on the inquiry with a predis- affectionately known as “Lemon Trees”.) position to find the department everything it ought to be, and a desire to restore public After an initial hearing in the Dunedin Res- confidence by declaring the charges against ident Magistrates court, George Barton was Government unfounded.” committed for trial in the Supreme Court. However, the Crown withdrew the charges (The Press [Christchurch] Tuesday, Novem- before the case commenced. Barton claimed ber 14, 1871) this was vindication of his earlier remarks. During the case in the magistrates court, the Barton left the ODT shortly after the case prosecution offered immunity-in-advance to concluded and went into legal practice in ODT staff who may have otherwise refused Dunedin (he had been trained and called to give self-incriminatory testimony. This to the bar in Australia prior to arriving in was claimed to be an abuse of power on the NZ). He authored two well-known books on part of the Governor, and Barton complained NZ legal matters, before returning to Aus- about this to the Colonial Secretary (in the tralia. Lemon served as General Manager of UK), but the response (if any) isn’t known. the Telegraph Department for over 25 years, Barton responded by counter-suing Charles before he retired. This fascinating chain of Lemon for a breach of the Electric Telegraph events shows that tension between politicians Act (1866), in that he (allegedly) supplied and the Press, related to the use of “modern” a copy of a sensitive ODT telegram to the information technology is not new! Also, Acting Minister of the Department, who had that politics (and basic human nature) hasn’t used it for political purposes. The Resi- changed much in over 140 years dent Magistrate refused to hear the case, on the grounds that the information was more David Wilton than six months old (i.e. statute of limita- tions). Barton appealed this to the Supreme Court but that was dismissed on a point of law. Both cases (i.e. against Barton and against Lemon) eventually petered out. The report of a parliamentary Select Committee Te Kooti’s Pa into the affair fairly naturally supported the Government position, i.e. that nothing un- Te Kooti’s Gunfighter Pa at Te Pononga toward had happened in any of the four in- cidents investigated. The views of the per- This article describes a search for a gun- ceived “anti-government” media were quite fighter pa site and it’s subsequent location different, however: and recording. The site is probably the de- fended pa used by Te Kooti and his follow- “Altogether the Committee seem to have ers in the hills above Tokaanu and Turangi, framed their verdict upon a famous and of- while engaging in military operations against 15

British forces in September-October 1869. It is a site of national importance, as it was the scene of a significant battle in the NZ Wars of the 1860s. As a number of soldiers were killed there, it is also possibly a military cemetery, as, in those days, soldiers killed in action were generally buried on the battle- field.

Background and History

During the late 1860s and early 1870s, much of colonial NZ lived in fear of a marauding Figure 7: Map showing Te Kooti campaign band of Maori guerrilla fighters under the sites around Turangi area (Wilson 1961 pp. leadership of Te Kooti Arikirangi Te Turuki; 32-33). known simply as Te Kooti. Te Kooti’s life and military exploits are well chronicled by but withdrew after capturing a number of numerous historians and authors probably horses, due to the impending arrival of the the most notable being the late Dame Ju- main body. “Te Kooti attempted to retrieve dith Binney, with her epic 1995 work enti- his failure to destroy the advance guard of the tled Redemption Songs: A Life of Te Kooti government forces by retiring into the hills Arikirangi Te Turuki. In 1867, Te Kooti was above Tokaanu, from where he could raid the accused of siding with Maori supposedly re- various contingents which now gathered to- belling against the British authorities and gether” (Wilson 1961 p.34). Wilson (1961) was exiled to the Chatham Islands. Te Kooti names Te Kooti’s defended locality in the always denied the accusations, and no doubt hills above Tokaanu as Te Pononga (see map this created considerable bitterness in his at- Fig 7). The existing main road (SH 47, built titude to colonial authorities. To cut a long c. 1970) between Tokaanu and Lake Rotoaira story short, Te Kooti and some followers hi- is known as Te Ponanga Saddle Road. jacked a small ship and escaped from the Chathams, returning to the Gisborne area. On 25th September, Te Kooti initiated two Here, they raised a small army and embarked engagements. The first was an attempted on a campaign of violence and destruction ambush on government forces commander against Pakeha settlers and the British au- Lieutenant Colonel McDonnell, as he rode thorities. between his camp at Poutu, at the eastern end of Lake Rotoaira, and the camp of Cap- According to Wilson (1961) War in the Tus- tain St George at Tokaanu. This was un- sock: “At the beginning of September [1869] successful. At about the same time, a gen- Te Kooti was in control of the Tokaanu area, eral attack was launched against St George supported by most of the local Maori popu- and his contingent at Tokaanu. This was also lation.” A variety of Government forces be- unsuccessful, and St George retaliated with gan to assemble in the area, and while they a counter-attack on Te Kooti’s main party. were still building up, there was an engage- Shortly afterwards, Te Kooti’s group with- ment in the Tauranga - Taupo area on 9th drew from the site in the hills above Tokaanu September. Te Kooti’s forces attacked an and moved to Papakai, and then eventually advance party of Hawkes Bay Maori under to Te Porere, where the final decisive engage- Henare Tomoana and had the upper hand, 16 ment of the campaign against Te Kooti took examination as the area is very difficult to place in early October 1869. reach on foot.”

The Survey

The Te Pononga gunfighter pa site was origi- nally recorded in 1965 on the basis of histori- cal evidence, and the appearance of trenches cut into a ridge in some oblique air photos, but had not been located on the ground. It was decided that the site was likely to be on Hinemihis track, which was then the main route between Tokaanu and Lake Rotoaira. This is a good tactical position for a force conducted mobile, guerrilla-type operations, as it offered the possibility of a delaying posi- tion if a hasty withdrawal was required. Any defensive position off the main route afforded the opportunity to be bypassed and hence cut off. A search was conducted, on probably one of the hottest days in January (or so it Figure 8: Sketch of Te Kooti in later life (ap- seemed). The bush was very thick re-growth, parently, he refused to have his photo taken). with almost impenetrable gorse and bracken in places. After about three hours, during According to Cowan (1922) a significant bat- which time we only progressed about 1.5 km tle was fought in and around the Te Pononga along the ridge, the pa site was located. The site, with several of Te Kooti’s men and two remains of defensive trenches, walls or - government soldiers being killed. As the nor- bankments and pits (including two bell pits mal practice in those days was to bury war for food or water storage) were found. dead where they fell, the site is possibly a war cemetery. The battle was also considered to be significant as it meant Te Kooti lost credibility as a military commander, which eroded his support base within Maoridom. The subsequent battle at Te Porere in Octo- ber resulted in a major military defeat for Te Kooti, and his threat was greatly diminished, although he and his followers continued to dominate the Ureweras area for a further decade. According to Newman (1988 p.5): “Some rifle pits used by Te Kooti when he occupied the forest around the Te Ponanga Saddle have not been located on the ground although they are recorded as N112/7.” The Figure 9: Remains of a bell pit, probably original site record (1965) states that the site used for water storage. was: “Observed in air search (photos taken) and located on vertical aerials but no ground After submitting an update to the site 17 record, we were encouraged to contact Judith Binney, to see if she was interested in the site - she certainly was! Her emailed comment was: “I spent a lot of time trying to work out where Te Ponanga was. Now I know why it Sni#ets of Hi

On 22nd June 1941 the German armies invaded the in three great columns - Army Group South headed for Kiev, Army Group Centre headed for Moscow and Army Group North headed for Leningrad. Leningrad was selected by Hitler as one of the three major targets for the following reasons: (a) it was politically and culturally significant: St Petersburg was the second biggest city in , founded by Czar Peter the Great in 1703 and the cap- ital of Russia for over 200 years (1713 to 1918). It was called Petrograd from 1914 to 1924 and Leningrad from 1924 to 1991. It is known as “the city of three revolutions” - the 1905 revolution, the February Revolution (in March 1917) and the October Revolution (in November 1917). It is home to numer- ous historical buildings (today it is a UN- ESCO World Heritage Site) including sev- Figure 10: Main defensive trench cut across eral palaces and The Hermitage (opened in the ridge at the eastern end of the site. 1764), the largest art museum in the world. (b) it was strategically significant: A German References: Cowan, J. (1922). The New army based at Leningrad could launch of- Zealand Wars: a History of the Maori Cam- fensives against the Moscow-Murmansk rail- paigns and the Pioneering Period, Govern- way. British and US “Lend-lease” equipment ment Printer, Wellington. Newman, M. and supplies, delivered by Royal Navy Arctic (1988). Archaeological Investigations in the convoys to Murmansk, flowed down this rail Vicinity of Lake Rotoaira and the Lower Ton- link to Soviet armies around Moscow and be- gariro River 1966-1971, NZ Historic Places yond. (c) the Soviet Baltic Fleet was based Trust, Wellington. Wilson, . (1961). War at the great fort of Kronstadt just west of in the Tussock: Te Kooti and the Battle at Leningrad. The capture of the city would Te Porere, National Historic Places Trust, ensure the destruction of the fleet. The big Wellington. guns of the fleet provided significant support to the Red Army during the siege. (d) it con- David Wilton tained numerous arms factories producing 11% of all Soviet industrial output in 1939. Many of the industrial plants were shipped to 18

Siberia as the Germans approached but sev- jected to artillery and air bombardments sev- eral important factories continued to func- eral times a day throughout the siege. (Over tion throughout the siege. 6,000 civilians were killed by bombing and shelling.) Within days, bombing had de- stroyed the warehouses containing most of the food supplies. Only boats, under con- stant aerial attack, kept supplies trickling in across .

It took only 12 weeks for the German and Finnish armies to surround the city. Leningrad was totally unprepared and local Communist Party leaders refused to ask for extra food supplies in case Stalin accused them of not being prepared. Almost im- Figure 11: German front lines at Shlissel- mediately, the 2.5 million civilians (includ- burg were about 20km apart. The track ing 400,000 children) and 200,000 Red Army was over 30km long (not straight!). The defenders ran short of food and Red Army managed to hold the Oranien- was introduced. By November the daily ra- baum pocket (protecting the naval base at tions had been reduced to: 500g for soldiers, Kronstadt) throughout the siege. 250g for workers and 125g for dependants (children, elderly and unemployed). The ra- On 27 June 1941, over a million civilians tion consisted almost entirely of bread mixed (mainly women and children) were enlisted (about 50%) with sawdust. (Rationing in to build fortifications and antitank defences Britain at its worst dropped to 320g per day along the outskirts of the city and they supplemented with home-grown vegetables.) constructed over 25,000km of trenches and many timber and concrete bunkers. When Army Group North arrived at Leningrad, they met stiff resistance from the Soviet re- inforcements manning these extensive forti- fications. It was easier for the Germans to continue eastwards past the city than to fight their way in. The last two trains that rolled out of Leningrad just before the rail line was cut were loaded with priceless paintings from The Hermitage. The museum buildings were bombed and hit by artillery fire during the siege and reopened in November 1945.

On 8 September 1941 the Germans reached the shores of Lake Ladoga on the far side of Leningrad and the Finnish army was ad- vancing from the north. The city now had no Figure 12: Residents flee from their apart- land link with the rest of Russia. It was the ments during bombing. first day of the 900-day siege the most de- Then came the coldest winter in decades with structive siege in history. The city was sub- temperatures falling to −30 ◦C. The Ger- 19 mans retreated from Moscow but they man- often conducted in complete silence. Sur- aged to retain the ring around Leningrad al- vivors recall watching others collapse silently though it was only 20km wide at Shlissel- in the street and walking on past them. As burg on the shores of Lake Ladoga. Elec- winter worsened, young mothers desperately tricity, gas, water supply and sewerage were seeking protein for their children, could be damaged by bombing and ceased to func- seen hacking pieces from the frozen corpses tion. Coal, oil and kerosene became unob- that littered the streets. Dead relatives were tainable. Furniture, floor boards and books concealed in basements so that their ration were burned for warmth. In the freezing tem- cards could continue to be used. peratures, people were starving. Every bird, rat, cat and dog in the city was eaten. Inge- By 26 November 1941, the ice on Lake nious recipes were concocted, including the Ladoga was 200mm thick and able to sup- stewing of cats intestines and seaweed soup. port trucks. A few supplies started to flow “Macaroni” was made from flax seed for cat- into the city along what was called the Road tle and “meat jelly” produced from boiling of Life. The ice road was used until 24 April bones and leather. Wallpaper paste (made 1942. As food came in, medical cases and from potatoes) was scraped off the walls and dependants (mostly children) went out. The eaten. Grass and weeds were cooked. lucky ones were packed into open trucks and then driven across the ice road, through bliz- zards and under constant bombing attacks. Most of the women and children had to walk in freezing conditions. The ice road did not connect with civilization and, at the far side, a 30km track was hacked through the wilder- ness - trucks often took a day to traverse the track alone. Over 1.4 million were evacuated and about 400,000 died during the evacua- tions.

Figure 13: An “official” corpse passes “unof- ficial” bodies on the way to the cemetery.

11,000 starved to death in November 1941. 52,000 died in December - about 1,600 per day. Over 4,000 died on Christmas day alone. These were the official deaths - those who had been reported and buried. But many died in apartments or in the streets and were never reported. The final official death total was 642,000 but it was actually probably over a million. Coffins ran out (wood was needed Figure 14: In the Russian winter, most travel for burning) and corpses were shrouded in was at night. Woman soldiers served as sheets or not at all. Extreme hunger and guards and navigation beacons. cold, as well as the sheer number of dead, robbed people of all emotion. Funerals were The deaths and evacuations eased the food 20 supply situation and rations increased. Al- in 1942 but died in July 1944 of tuberculosis. though there were to be two more siege win- Tanya and her diary have become an iconic ters, conditions were never again as bad as image of the victims of the siege and her di- that first winter. In the summers of 1942 and ary can be seen at the Museum of Leningrad 1943, every patch of open ground sprouted History in St. Petersburg. vegetables to supplement the rations. On 27 January 1944, the Red Army finally forced the Germans away from Leningrad and the siege was over. Visitors to modern St Pe- tersburg can tour the Piskariovskoye Memo- rial Cemetery where about 500,000 people are buried in 186 mass graves. It is a lo- cal custom to place bread as well as flowers on the graves.

The picture shows Tatyana (Tanya) Savicheva and the last nine pages of her diary. The photo of Tanya was taken in 1936 when she was 6 years old. The diary was a small notebook with one entry per page. The English translation follows:

First row: 1. Zhenya (sister) died on 28th Dec. at 12.00 noon 1941 Figure 15: Rations for one day. 2. Grandma died on 25th Jan. at 3pm 1942 Tanya Savicheva was 11 when the siege 3. Leka (brother Leonid) died on 17th March started and she kept a diary. Her father at 5am 1942 had died when she was 6 and she lived 4. Uncle Vasya died on 13th Apr. at 2am in Leningrad with her mother, her grand- 1942 mother, two uncles, a brother and two sis- 5. Uncle Lesha died on 10th May at 4pm ters. Tanya was over 10 years younger than 1942 the others. Her mother, brother and sisters worked producing military supplies and her Second row: uncles served in the anti-aircraft defences. 1. Mama died on 13th May at 7.30am 1942 Tanya helped to dig trenches and put out 2. Savichevs died incendiary bombs. One day her sister Nina 3. Everyone died went to work and never came back. The fam- 4. Only Tanya is left. ily assumed that she was dead but Nina had been evacuated across Lake Ladoga and sur- vived the war. Tanya herself was evacuated Chris Scogings 21

Wood Pigeon

Figure 16: Wood Pigeon by Dave Wilton.