The First Welsh Footstep in Patagonia

The Primitive Location of Port Madryn

Versión en castellano

Autor: Fernando Coronato coronato@cenpat.edu.ar

This work has been published in the Welsh History Review, Vol.18, N°4, December 1997.

I wish to thank this Review, its Editor, Prof. Kenneth Morgan, and the University of Wales Press for granting permission to reproduce this contribution on Internet. www.wales.ac.uk/press

The arrival of Welsh settlers in Patagonia took place within the frame of the great European immigration, which colonized Argentina in the second half of the nineteenth century.

The main interest of this venture is the fact that this settlement was the first successful attempt in colonizing Argentine Patagonia, south of Río Negro (41°S).

In 1843, Chile occupied the Strait of Magellan. Ten years earlier, Great Britain removed the Argentine authorities from the Falklands. French and North American seal-hunters frequented the Patagonian coasts freely. Only the Argentine sailor Luis Piedra Buena exhibited with zeal the banner of his country at Isla Pavón, in Santa Cruz (Fig.1-a).

The establishment of a colony in Patagonia was well-suited to the strategy of the government of Buenos Aires in establishing its authority in the region. Also, the enterprise was important for a group of Welsh nationalists searching for a vacant land in which to develop their culture unfettered1; a purely Welsh settlement overseas where all the characteristics of the people could be preserved2 or, in other words, a refuge from cultural and economic oppression in Wales3.

The details of preparations and installation are well known. Reading the general chronicles written by Abraham Matthews and Lewis Jones is enough to realize the extent of early difficulties.

The first group of settlers, about 150 people, sailed non-stop from Liverpool to Patagonia aboard the tea-clipper Mimosa. On 28 July 1865 they landed in New Bay, the western extremity of New Gulf. Soon, that point was to be known as Port Madryn (Fig.1-b).

The subsequent story of the Colony is known almost in every detail and has often been told4. Nevertheless, the place where the Welsh settlers built their first dwellings has not been unequivocally established. The colonists move away from the coastal settlements they has originally made so that today only their names remain, such as Port Madryn which displays little Welsh except for a monument on the seafront in concrete and bronze which marks its foundation5.

So, the first location of Port Madryn is open to discussion. Some people believe that the first encampment was in the northern part of the city, at the bottom of the hill upon which the Chalet Pujol rises today (Fig.2-a). Formerly, heavy rainfalls used to form some ponds in that point (Laguna de Derbes). The local board of historical studies endorses this hypothesis6. However, for the municipal administration, the southern district of the city is called Desembarco, according to a widespread tradition.

Oral tradition has always said that the first colonists sheltered in "the caves." Obviously the caverns of Punta Cuevas, which are submerged by the high-tide twice a day, could not serve as refuge. On the other hand, very near, on the west side of Punta Cuevas, there is a row of artificial excavations, situated above the high-tide level (Fig.2-b). It is likely that these pits could have been primitive shelters. Their shape is regular and their size is that of a little room. There are five of them, quite well preserved; two others are half-destroyed and there are just vestiges of an other one. Along the stretch of coast following immediately, over 45 m, there are several marks of the same man-made work. Owing to its position, the sea eroded the latter space to a greater extent. One can just suppose the existence of eight or nine more hollows. Thus, the total number of "dwellings" would have been fifteen or sixteen. (Fig.3)(Table 1).

Despite these facts, there is no certainty about the ruins of Punta Cuevas. We do not know whether the colonists spent their first Patagonian weeks in these caves, or that they are contemporaneous with their arrival. Since the witnesses of that event must have experienced the moment so intensely, it is hardly surprising that they did not stop to write down the details. Due to this lack of precision, the available information was not considered, allowing some local searchers to suppose that the original location of the Welsh settlement was elsewhere. The fact that most of what has been written about the settlement was published in Welsh7, also helped to mislead, since Spanish translation is recent and still incomplete. Because of this, it seems useful to review the available information on the subject. The aim is to verify the oral tradition referring to the landing and settlement of the first Welsh colonists in Punta Cuevas, and to confirm that the existing remains belong to this settlement.

2. Documentary sources.

Seven diaries referring to the arrival of the Mimosa are known. Edwyn Roberts and Lewis Jones were the first to arrive, in June 1865, to make preparations for the settlement. Both mention the building of stone-made constructions, which were finished with timber. Neither Edwyn Roberts8 nor Lewis Jones9 mentions the existence of any excavation when they first arrived in Port Madryn; however they do mention the ruins of the Old Fortress when they went to TreRawson. The row of hollows at Punta Cuevas is something that attracts attention even nowadays. Certainly, it could not have been unnoticed to those who spent several weeks in that bay, so one can presume that the hollows did not exist before they arrived.

In 1854, Henry Libanus Jones tells that there were remains of whalers' boilers in the western extremity of the New Gulf but he does not mention the existence of any man-made construction10. Owing to these boilers, one of the first names of Port Madryn was "Pot Harbour"11 and they are mentioned until as late as 188612.

Among the travelers of the Mimosa, there are the records of Abraham Matthews13, Thomas Jones Glan Camwy14 and Lewis Humphreys15. All of them talk about a row of huts. The travel diaries of Hugh Hughes Cadfan and Joseph Seth Jones stop on the arrival date. The latter, quoting some parts of Lewis Jones' welcome speech16, refers to sixteen "houses".

Again in 1867, we are clearly told that the settlers sheltered in the pits dug into the rock. Yet we do not know if these cuttings were built then or if they had already been made17.

In the following twenty years, a short series of documents describing the area appeared (maps, surveys, reports). Although they do not mention the location of the settlement in Punta Cuevas directly, they furnish several indications that support this position and perhaps the continuance of some sporadic activity at this point of the coast 18. This series of documents starts with the 1866 report of the land-surveyor Julio Díaz19, who plotted lots and measured off the farms in the Chubut Valley. Captain Dumistown's report, of the British gunboat Cracker, gave a complete description of the Colony in 1871 20. In 1876 the Captain Fairfax's report, HMS Volage, presents a chart of the western extremity of New Gulf21. There is another map of the same area, signed by Comm. Lasserre in 1881 22, when the Argentine government prepared the installation of a port in that area. The series ends in 1884 with the Eng. E. J. Williams' work, who made the map of the lands belonging to the Central Chubut Railway Company23.

In the mid-1880s, two documents speak directly of the Welsh settlement in Punta Cuevas. They are two narrations of exploratory voyages to the Valdés peninsula, written by Thomas B. Phillips in 1885 24 and by Eng. Llwyd Ap Iwan in 1886 25. Although the authors were not witnesses of the 1865 landing, they were both notable colonists and were undoubtedly acquainted with its history prior to their arrival. Also, one must consider the possibility that some of the travellers who arrived in 1865 could have taken part in these exploration trips, and in that case Phillips and/or Ap Iwan, would have received first-hand information.

This kind of information is only found in a 1910 publication. In that year, a witness of the first landing, Richard Berwyn, wrote some complementary annotations to Lewis Humphreys' diary, published by Y Drafod when the Humphreys died.26

Later historical works, when referring specifically to the 1865 landing, are based on the sources cited above or quote the oral tradition. However, in 1931 the Port Madryn local newspaper briefly commented that vestiges of "the first Welsh dwellings" were discovered at Punta Cuevas27. Ithel Berwyn 28 and Evan Thomas29 placed the landing in Punta Cuevas, while other authors mention the shelters dug into the rock: Howel Davies30, Bryn Williams31, E. Bowen32, Glyn Williams33. Juan Meisen's 1984 report34, which deals with the subject of the landing in a secondary way, is the only one to introduce elements of doubt.

A careful comparison between Lewis Jones' diary35 (henceforth, in this section [LJ]) and Edwyn Roberts' writings36, 37 (henceforth [MS1] and [MS2]), allows us to order the events and to deduce details about the first encampment location. Both are complementary works, since Jones gives exact dates while Roberts retains the main facts but not the dates, surely because he wrote long afterwards.

Coming from Patagones aboard the schooner Juno, we sailed a few days [MS1] (two days)[MS2] along the coast, looking for an appropriate landing site, which was found at the top of the bay, beside the white rock [MS1]. They disembarked on 14 June at noon and the sheep and horses were then downloaded [LJ]. On 16 June the cows and the timber were put ashore [LJ]. Although some corrals had been built since the first day, the losing of animals is pointed out more than once until 20 June. On 17 June we could use the cart to carry tosca blocks [LJ]. On 19 June the foundation of the first house is placed [MS1][LJ]. On 20 June it rains heavily and only on 21 June, after an entire week looking for water in vain, luckily we found water, although it is not wholly fresh [LJ]. Roberts does not point out on what date the animals are moved about 4 miles up the bay, to the site which is nowadays called Port Madryn, where ponds formed by rainwater [MS1]. Thus, the building of the first constructions had begun before finding water, that is to say, that the location of the encampment was not related to that of the water. This water-point was situated 4 miles northward and only the livestock were moved there. The ship was taken nearer the new site and was unloaded, by piling up the burden upon a dune, above the line of high-tide [MS1]. A corral is built for the sheep in that place.

On the same day, 21 June, we began building the dwellings with timber instead of tosca, because this is too slow [LJ]. It is not explained whether the dwellings were built near the first ones or next to the water-point. It is likely enough that at this point some kind of shelter was raised to be used by a shepherd or watchman. Building materials had been unloaded at both places. On 25 June, the wall of tosca, built with such difficulty, was knocked down by the wind [LJ]. The encampment continued to be built as a row of houses with the timber brought from Patagones [MS1]; some boards available from a ship wreck were also employed [LJ]. Later, when the ponds dried out, Edwyn Roberts decided to dig a well in search of water..I went down 8 yards until finding it, but it was salt water... [MS1].

Wherever the site of the encampment, that is where the colonists of the Mimosa landed a few weeks later. From the writings of these people, we can get some indications referring to the location of the first dwellings of Port Madryn. There were a few wooden huts on the seashore, built as a shelter for women and children38. They were 41 women and 61 youngsters 39.

Thomas Jones, Glan Camwy, talks about a store dug into the tosca, roofed with rushes from the seashore and a sack instead of a door. He adds that the houses were a few planks put vertically, forming a row of about twenty rods...they were just to sleep and to shelter40.

Lewis Humphreys' diary adds an important detail: On July 31, at 4 p.m. William Jones' baby was buried in the hills of the southern side of the bay41.

The circumstances in which some people were missing the very day of the arrival give us clues:

they called me to go with some others for water, to where the railway station is now (1902); David Williams came with us and we went chatting on our way to the water, which was muddy... since the animals had trampled inside...Soon it became late and as we began the way back, David Williams was not there. We had to return to the encampment to give the news42.

On the other hand, ...some women and girls who had gone for water...we sought for them, we lighted fires, we cried out loudly, but they appeared only the next morning, near the water-point, tired and numb with cold43.

These two accounts allow us to suppose that the encampment was not located near the water-point (that is, it was not near the site that was to be called Laguna de Derbes). One had to "go for water" and the way took some time ("we went chatting on our way," "we began the way back," expressions that denote a considerable distance, long enough as to avoid visual or hearing contact). Although water may be carried from any distance, people do not behave in the same way with the dead, and usually they are buried not far from the settlement. So, one may suppose that the young girl and the others who died before leaving for the Chubut Valley (four children and a woman) were buried near the place where they encamped44, that is to say, in Humphreys' words, in the hills of the southern side of the bay45. The fact that the encampment was at a considerable distance from the water-point is confirmed by Thomas Jones, Glan Camwy's description: As there is about three miles to carry the water, the committee decided that two barrels a day could be employed in food ...and that salt water of the well must be used as a complement46.

The distance of three miles is the same as is pointed out by Edwyn Roberts, who located the water-point at the place where Port Madryn is now47, or, more exactly, where the Railway station is now48 (that is Laguna de Derbes). It is likely that the moving of the company to the Chubut Valley was achieved by the beginning of September 1865. The last death recorded in Port Madryn was on 22 August49, while the ship in which most of the women and children were carried got under way for Rawson about 5 September50. Presumably, not many things were left in New Bay. The store of tosca was still in use for at least three more years51. However, it is likely that most of the wooden huts, raised some weeks before, were disassembled to be used again in TreRawson. Building materials were scarce and it seems reasonable to think they were not left abandoned. Yet the stone buildings, remained unavoidably in New Bay.

The hollows of Punta Cuevas are mentioned again two years later. In the winter of 1867, all the settlers came back to Port Madryn ready to abandon Patagonia. They met an Indian tribe in their winter camp on the seashore. John D. Evans tells us that the cacique (Galats) was astonished to see the primitive and precarious way some colonists lived, sheltered in the carvings on the tosca and in the caves52. On his own, Thomas Jones Glan Camwy remembers those days saying: ...after arriving there, every one made some kind of hut to his family had some shelter during some time, because we did not know when it would arrive a ship which take us out from there. Two people dug a room in the hillside, which can be seen nowadays ...others took planks from the old wreck...53.

It is highly probable that the colonists had occupied the excavations made two years earlier. Maybe they had to restore them partially or readapt them to new situations. The fact that two men dug a room in the hillside must not allow us to lose sight of the fact that the rock had been worked in 1865 already, as was clearly stated by Edwyn Roberts54 and Lewis Jones55.

In the midst of the problems of that decisive winter, Lewis Jones remembers that once the uproar was calmed down, we celebrated the Landing Day in the caves and the beach of Port Madryn56.

When the settlers returned to the Chubut Valley, one of them described a sort of dwelling which is exactly the same as the shelters at Punta Cuevas:

Our precarious dwelling was on the side of a hill, and to have a room, we chose cutting the hillside; doing so, we had three natural walls. We put some branches and shrubs as a roof. This roof was at the same level as the hillside; the room was large and we used it as a kitchen57.

This description makes it evident that the colonists knew about this kind of construction. Another proof of this is given by Abraham Matthews when he states that some of the first houses built in TreRawson in 1865 were dug into the Old Fortress rampart58.

The next indirect reference that suggests the passage of Welsh Colonists through Punta Cuevas is found in 1881. In that year, Commander Augusto Lasserre produced a detailed map of the western extremity of New Gulf 59, when the Argentine government was planning a port in the area. He referred to Punta Cuevas as Punta Galenses (Welshmen’s Point) and called the adjacent inlet, nowadays called Curva del Indio, Bahía Galenses (Welshmen’s Bay) (Fig.2-b). What could have induced this navy officer to give those names to a particular stretch of coast, if it were not the remains of the Welsh presence fourteen or sixteen years earlier? Such names would be unexplainable if the landing site and the encampment had been in Laguna de Derbes, as is often argued. Also in 1881, the existence of remains of a settlement in Punta Cuevas was pointed out by C. Lambert, an excursionist of an English ship. He described the buildings but supposed that they were more likely the work of sealers. Yet with the path starting there and he meeting three colonists of TreRawson, the participation of Welsh settlers in the buildings should have occurred to him. Lambert's description tells us that: ...we passed the afternoon in rambling about the shore and examining some curious cabins cut out of what can hardly be called rock...One of them had a door, window, fireplace, and a cupboard, and a bed-place cut into the stone..60.

On the other hand, on a British chart of the New Gulf, surveyed in 1876 61, the name Cave Bluff first appears and a wreck is shown at the bottom of the inlet. The sole man-made geographical element drawn on this chart, is the road "to Chupat," starting from Curva del Indio and striking straight southward. (Fig.2). The features of this road coincide with Edwyn Roberts' chronicle since (while waiting for the arrival) I started tracing the road to Chubut by clearing shrubs, tracing a straight way. I had done about 5 miles when the first company arrived62. This task continued after the arrival, since during the time the colonists spent in Port Madryn we are told that ...Some others began to grub shrubs to make the way between Port Madryn and the Chubut Valley63; or, there were some men making a road toward the valley, clearing the bushes...and there are several miles of this road that remain practicable until today 64. The works continued, since the land surveyor Julio V. Díaz reported in September 1865: they have cleared up about the third part of the distance between New Bay and the Chubut 65 (that is, about 13 miles). From Capt. Dumistown's 1871 report, it follows that the road had kept its straightness: There are some straight roads such as the one of New Gulf, which stretches over 12 miles66. Until as late as 1884 the same "cart path" is the sole road appearing in the area of Port Madryn according to the map surveyed by Eng. Edward J. Williams67.

Throughout nineteen years none of the references to this road suggests any lengthening further north, as one would expect if the first encampment had been in Laguna de Derbes. In 1865, Edwyn Roberts described a landscape that is consistent with the one found southward from Punta Cuevas, following the way mentioned above68. Although he overestimated the distances by almost 30 per cent in relation to the actual ones69, the ratios between them coincide accurately with the reality, which would not be so if one started in Laguna de Derbes. (Table 2).

Another argument that favours the choice of Punta Cuevas as the first encampment follows from the search for a safe and easily defensible site. Expressions showing the fear of the Indians are abundant in Edwyn Roberts' writings70 (his experience in North America and the stories told in Patagones must have filled him with respect for them). The bluff of Punta Cuevas is a good watch-point over the neighbouring countryside, the south and the west. These are the only vulnerable flanks because the eastern and the northern sides are protected by the sea. These defensive conditions are far more advantageous than those of Laguna de Derbes.

The hypothesis of the landing near Laguna de Derbes emerged in the mid-1970s. In Port Madryn, those were years of overwhelming population growth due to internal migration. Thus, the links with the local past, by then rather weakened, could be broken with almost no reaction.

Meisen's work on the short and circumstantial passage by New Bay in 1863 by Sir Love Jones Parry of Madryn and Lewis Jones was well based and had some academic support71. Nevertheless, the hypothesis of the landing at Laguna de Derbes, included in that work as a complementary subject, is not treated as deeply as the core of the work. When referring to the arrival of the Mimosa in July 1865, Meisen rejects the landing in Punta Cuevas, merely because he considers that it is an unsuitable point. Yet the Argentine Navy, in describing the coasts of New Gulf, asserts the opposite72. Documents published after Meisen's work furnished information that would surely have led this author to revise his opinion73. Despite the weakness of this hypothesis, it has been widespread in recent years, by means of lectures, printing and tourist information brochures74.

The two documents of the mid-1880s that locate the first encampment in Punta Cuevas in an explicit way, have several points in common. Both are narrations of exploratory voyages to the Valdés Peninsula, during which the area of Port Madryn was chosen to stay at night. Both were written by prominent colonists and, although they had arrived a short time before in Patagonia, one can suppose they wrote with some accuracy. The first narration was written in 1885 by Thomas B. Phillips, who states:

This angle of the bay had been named by the colonists Port Madryn...this was the place where the explorers determined to bivouac and while some drove the horses further up the valley...other removed the baggage to some caverns, which in part had been excavated by the first colonists when they landed, to serve as habitation. Many of these caves bore witness to the ingenuity of their excavators in the shape of ornamental mouldings which had been carved in the clay... Some strolled to a remarkable cave which, as the tide was out, they were enabled to enter. They found the floor covered with shells and the walls with inscriptions, being mostly the names of the officers of the British gunboat Cracker, which some previous years had been sent to visit the Colony 75.

The second document is the narration of the Llwyd Ap Iwan's trip in May 1886. Michael D. Jones' son and his companions made a stop to stay the night at the hangars of Puerto Roca, the one installed by the Argentine government.

Puerto Roca is the name of this haven in the New Gulf; after passing a point that advances into the sea, in front of the salt water well, here we are in Port Madryn, a better and more protected haven. Along the western flank of the mentioned point, on the Port Madryn side, there are remains of the old colonists' houses, where they lived while waiting to leave, twenty years ago. The houses were cut out of the soft and clayey rock (tosca), it was a row of hollows dug into it and in that place the cart road to TreRawson starts. There is a wreck at the bottom on the sand. I think it was there before the arrival of the first settlers; and further upward, on the top of the beach, there are big old boilers, it seems there were sealers ships that melted the oil on land...After walking four miles we arrived to the Great Valley, as the old colonists named a brook, or basin, which runs from a few miles westward Port Madryn. We went to the outlet in the sea, but the old stream was then dry. The town of Port Madryn raises in this place since this is where the railway to the Chubut Valley starts76 (Fig. 2-a).

After reading these two documents almost no doubt remains. Yet it is surprising that Ap Iwan mentions that it was the place where the settlers camped in 1867 and says nothing about the arrival in 1865. Perhaps it was obvious to him that he was dealing with the same place. The same happens with John D. Evans, who refers to the hollows in 1867 but not in 1865, when he was only three years old77. In 1867, Lewis Jones talked about the caves and the beach of Port Madryn in a plain and spontaneous way as if speaking about something already known and not about something newly built78.

Despite being vague in the dating of the caves, Ap Iwan's document furnishes information which overcome this omission. Three elements, the well, the boilers and the wreck, are mentioned as being close to the ruins of the encampment. Such elements had already been mentioned as close to the 1865 settlement. The first element is the well dug by Edwyn Roberts in June 1865 79; it is also mentioned by Thomas Jones, Glan Camwy80 and John D. Evans81 in the same year. The latter’s narration allows us to assume that the well was southward, on the road to TreRawson, not far from the encampment. The second element, the boilers, was cited as being there in 1865 82 or as early as in 1854 83. Finally, the wreck, the third element, is mentioned in 1865 by Lewis Jones84. The probability that this triple coincidence could occur somewhere else in New Bay, is very low (especially the well), so one may conclude that the place of the 1865 encampment and the place described by Ap Iwan are, in fact, one and the same.

In turn, Thomas B. Phillips does not hesitate to link the caves to the first landing. His mention to a remarkable cave (obviously natural as the tide went in) next to the hollows, is consistent with the geography of Punta Cuevas.

The only witness of the arrival of the Mimosa who published any reference to the location of the landing site was Richard J. Berwyn. When Revd. Lewis Humphreys died in 1910, his travel diary was published by Y Drafod, then directed by R. Berwyn. He considered it useful to add some details originally omitted. Thanks to these annotations we know, for instance, that the day of the arrival...

the flag with the red dragon in its middle floated on the top of the hill that dominates the caves, and in the eastern corner of the inlet, between the beach and the dunes, there was a store ...and a row of huts on timber half raised, which were finished in the following days85.

Thus, Berwyn explicitly confirms the deductions about the location of the first encampment discussed above, as well as the references of the two mid-1880s documents. Despite being succinct, this witness is especially trustworthy since Berwyn was very accurate in all his notes. He was the secretary of the committee during nine years and kept the demographic register during the first ten years. In 1867, he was designated to register the traffic of ships in Rawson and Port Madryn. In 1871 he was named Post Administrant and, in 1875, delegate of the Migrations Bureau. In 1880 the National Weather Service entrusted him the first systematic meteorological records in Patagonia. Besides being the "official recorder" of the Colony in some sense, his duties in the longshore movement and in the Post, took him often to Madryn in those early years86. Therefore, Berwyn was an eye-witness of the landing of the Mimosa who was particularly well-placed to testify about its location.

In agreement with Ap Iwan's description, the ruins of the excavations, still visible today, are disposed in a row along the coast, on the western side of Punta Cuevas87. This arrangement coincides with the 1865 information given by Edwyn Roberts88, Richard Berwyn89 and Thomas Jones, Glan Camwy90.

The back-bottom edge of an excavation (n°16?) is still noticeable in the southern extremity of this eroded span of coast, 80 m away from cave n°1. The floor level of these shelf-like remains, differs only 5 cm in height to that of cave n°1. (Fig. 3-a). All in all, the existing remains suggest a systematic construction as described by Edwyn Roberts in 1865 96, rather than an unplanned or isolated one, as may be supposed from Thomas Jones, Glan Camwy when he tells us that in 1867 ..two people dug a room in the hillside...others took planks from the wreck97.

In 1994, because of the enlargement of one of the modern constructions mentioned above, about 100 glazed ceramic bottles were found in the site. They were disposed properly, piled horizontally, like in a store-place and not in a waste-place. These bottles are glazed outside and inside and labeled "Price. Bristol". Their capacity is a British pint (0.55 l). According to some specialists, they were manufactured about 1880-1890 and used to contain beer98. This finding seems to confirm that the site of Punta Cuevas was frequented and/or occupied until the beginning of the railway construction in 1886. This agrees with some of the evidence further detailed99, 100.

As a result of some minor road works, the remains of a woman were discovered in Punta Cuevas in September 1995 101. They were 180 m SSE from the row of hollows (Fig.2-b ). The body had been buried in a wooden coffin, at about 50 cm depth. A gold wedding ring and a button made of mother-of-pearl, were the only accessories found102. At present, the analysis of the remains shows that they belonged to a middle-aged Caucasian woman103 and that the coffin was made of Scots pine (Pinus sylvestris)104. Since this kind of timber may be used in ship construction105, it is supposed that the coffin might have been made with planks obtained from the wreck on the adjacent inlet. The use of this timber in building the 1865 encampment is pointed out by Lewis Jones106.

It is likely enough that this grave was that of the sole adult person who died during the time spent by the 1865 company in Port Madryn. It was Mrs. Catherine Davies, of Llandrillo, who died when she was 38 years old107. Studies are being carried out by the Centro Nacional Patagónico, in Port Madryn. If the identity of these remains is confirmed, it would be a conclusive evidence of the location of the first Welsh settlement in Patagonia.

8. Hypothetic reconstruction of the first settlement sequence

Based on the information exposed above, it is possible to describe the sequence of the first settlement.

In June 1865, Lewis Jones and Edwyn Robert chose Punta Cuevas because it was an easily defensible point, which offered soft rock good enough for building. The site was close to a sheltered beach. Yet they ignored the lack of nearby fresh water. In the search for an appropriate place, the presence of a wreck, furnishing supplementary timber, may also have been considered.

Another possibility to keep in mind is that in Patagones, Jones and Roberts might have learned the feasibility of digging shelters in the rock. The men they hired in Patagones possibly suggested the idea of such shelters, like the Cueva de los Maragatos in that town. The excavation of the hollows began two or three days after landing, that is, before finding fresh water. The stones removed from the hollows were employed in building the storehouse, to derive double benefit from a single effort. Three men worked during a week in stone cutting and in moving the blocks from that point to the beach, about 100 or 200 m apart. Perhaps these were the unusual bricks found in 1931 108. A heavy rainfall ran on the Madryn basin, filling the ponds near the outlet, 4.4 km northward of Punta Cuevas. The fresh water was discovered the next day, because the fluttering of birds. Yet it was too late to think in raising new dwellings there, and that is why only the animals were moved close to the water-point. A corral was built up there, and perhaps a precarious shelter too; the ship was taken nearer the new site and was unloaded 109. At the first point, at the top of the bay, beside the white rock, planks and zinc sheets had been unloaded. They would be employed in building and roofing the huts and the store. Also, the wreck was employed for that purpose.

Edwyn Roberts and Lewis Jones probably knew that in 1854, Henry Libanus Jones mentioned the existence of this water-point110. Yet when they sailed two days111 along the coast, looking for an appropriate landing site112, they did not find any pond. Surely it had dried up by then, as it was again a few weeks after being found: the water in the pond dried out soon, and even though we looked for it everywhere we could not find it 113 (as faraway as one-day-horse-riding)114. Luckily, before the arrival of the company, the ponds would fill again115.

When the Mimosa arrived, the landing of people took place at the encampment in Punta Cuevas, or not very far away from that point. People could have landed on either of the two beaches situated on both sides of the point. Both are suitable to be employed to disembark116. According to Thomas Jones, Glan Camwy117 the travellers of the Mimosa landed in the western side of the point, that is, in "Bahía Galenses". The storehouse and the well made by Edwyn Roberts, must have been in the neighbouring dunes southward, not far from the grave discovered in 1995. The row of huts, at least in part, is the range of hollows visible now. All this leads one to think that, though the dwellings were dug into the rock, their front facing the sea was made of planks placed vertically. From the sea they did not look like shelters but like a range of planks. 118. The roof of zinc sheets or bushes could be easily supported by the frontal planks and the opposite rocky wall. The best preserved excavation (the first one from north to south) still conserves the indentation of the frontal wall carved in the rock, as well as remnants of a wooden lattice119.

Not all the cottages were finished when the first company arrived, but they were...half-raised and were finished within the following days120. The carpenters were eager to get the materials to building houses that they had brought from Liverpool..121. Richard Berwyn agreed to this when he wrote that the carpenters finished the sixteen houses which Roberts and his helpers had began122. Since the raw of huts was about 100 m long 123 and the existing ruins stretch over almost 80 m, one may deduce that there were few demonstrable shelters, if any. Anyway, the covered surface prepared before the landing was enough to lodge all the people (Table 3), even more if the store is computed, since it is known that a family was lodged in it124.

In 1881, Lambert's description denotes the well preserved state of some of the cottages125. This fact leads one to think that at least one shelter was still in temporary use. Thomas Jones, Glan Camwy's witness is clear on this point: ...it is a couple of rooms that... from 1865 to 1887, served as habitation to the envoys of the authorities of the Colony, to receive the mail, shipments or passengers which could arrive126. In 1932, L. Deodat visited the cave which had furniture carved on its walls. He observed two small vaulted fireplaces almost at the level of the floor and in front of that, on the bottom wall a sort of cupboard, two thresholds at both sides and...three rude stairs...127. According to this description, it should be the excavation n°2, which has carved walls and the vestiges of the stairs.

There is, as we have seen, enough evidence to support the hypothesis that the first Welsh settlement in Patagonia was located in Punta Cuevas. However, it is surprising that a comparatively recent historical event, quite well documented, has been veiled by uncertainties. First, until the 1880s, the existing sources such as Ap Iwan's and Phillips' narrations and also the Lasserre map were clear enough about the location of the first encampment. From 1886, due to the railway construction and the appearance of the town of Madryn beside the Laguna de Derbes, the longshore activity in New Bay was no longer based on Punta Cuevas and moved to the new port. The site of the 1865-1867 settlement was no longer frequented and became forgotten. The members of the first arrival lived in the Chubut Valley and between 1890 and 1920 their generation disappeared. As early as in 1889, not even thirty people remainded to remember the details of the foundation of the Colony128. The flood that devastated the Chubut Valley in 1899, destroyed a great deal of the written memory of this first generation. For instance, most of Richard Berwyn's records were then lost129, as well as many of Lewis Jones' papers130. The records were lost in the flood...and then there was nobody afterwards131. The writers on the founding of the Colony overcame this gap by relying mainly on oral evidence. They got most of the information from other people and some of them have no connection with the first pioneers 132. The stories passed down from one generation...to another, have diminished in accuracy in the re-telling133. J. Taylor also regrets the inaccuracy of oral testimonies, that has been somewhat confused along the line..as if a certain element of Celtic magic had been injected into the sequence of chronicles 134. Secondly, there are some signs that allow us to suppose that by the first years of this century the ruins visible today in Punta Cuevas were totally or partially covered by sand. In 1931, the information given by the weekly Golfo Nuevo tells us that works had to be done to "put away the earth" and to "throw the sand into the sea" to discover the vestiges135. On the other hand, the natural caverns existing in Punta Cuevas also exist in the neighbouring places (Points Arco, Este and Loma). So, one may suppose that the name "Cave Bluff" first appeared in 1876 136, does not refer to the caverns which are a repeated feature of the landscape, but to the artificial caves, which are found only in Punta Cuevas. The Spanish word for "cave" (cueva) does not include the concept of human building, and this meaning was lost when the name of the point was translated into Spanish (cave > cueva). Thus, if the excavations were hidden by the turn of the century, it is logical to think that the name "Punta Cuevas" had been linked to the sole caves that were evident at that moment, that is to say, the natural caverns. Hence emerged the absurd, but widespread, idea that the colonists of the Mimosa sheltered in the natural caverns of Punta Cuevas. Some authors, like C. Borgialli137, L. Amaya138, J. Dennler139, and R. Gorráiz Beloqui140, transmit this view without considering that the high-tide invades the place twice a day. On the other hand, Thomas Jones, Glan Camwy wrote the following:

Because many things are told referring to the caves located in the other side of the hill, facing (peninsula) Valdes, as that the first colonists sheltered there when they landed, the people who tell this should think twice before continuing doing so. That place is a suitable one for fishes since during the high-tide there is as much as one metre of water. Surely it is not an appropriate place for human beings141.

Certainly, one has to agree that the confusion is not surprising since a single word ("cuevas") is employed to designate three different things, located in the same point: a) a geographic accident, b) the natural caverns and c) a human construction.

Finally, the last contribution to confusion was introduced by the hypothesis asserting that the first encampment was located near Laguna de Derbes142. Although the 1865 settlers frequented that water-point, their encampment was 4 km away143. Probably Meisen, finding illogical the idea of the shelter in the natural caverns of Punta Cuevas, imagined an alternative location to the landing of the Mimosa, and in doing so took some elements of the arrival of the Vesta in 1886. In that year the newcomers settled beside Laguna de Derbes, but they spent the first days in Punta Cuevas, and we know that a baby was born there144.

All documentary sources agree in indicating Punta Cuevas as the point where the Welsh first settled in 1865, confirming in this way the oral tradition. Once the origin of the ruins of Punta Cuevas is known, it becomes easier to plan further research in order to complete our knowledge. It would be interesting to know where the well was situated, to find the message that Edwyn Roberts put under the foundation of the storehouse, and to locate the hollow where bread was baked.

The certainty that the remains of Punta Cuevas are the first footsteps made by the Welsh Colonists in Patagonia compels us to preserve and to revalue this site. Punta Cuevas, because of the beauty of the scenery and its easy access, is a highly vulnerable place, which must be carefully managed. Although the existing ruins are rather humble, they mark a valuable historical site because they reveal something of very beginnings of the first successful attempt in colonizing Patagonia.

I am very grateful to Mr. Ellis Roberts (Trelew) for lending me the Edwyn Roberts' writings and for helping in the survey of the caves. I am indebted to Mrs. Tegai Roberts (Gaiman), for her invaluable documentary research. Mr. Tomos Roberts (Bangor) and Mrs. Ana Beeskow (Madryn) contributed also with some clue documents. Also I wish to acknowledge the valuable advice from Elvey Mac Donald (Aberystwyth). Mr. Guillermo Defossé and Mrs. Patricia Hopkins (Madryn) revised the English version of this manuscript. My sincere thanks are due to all these persons.

Estimation of the total number of excavations, based on field measurements.

existing excav. = 7 stretching over = 35 m

disappeared ex. = x stretching over = 44 m

x = 8,6 (i.e. 8 or 9 disappeared excavations)

Actual topography compared to Edwyn Roberts' description⁶⁸.

Ratio between distance and relief (distances in miles).

(a) gentle slope

(b) plateau

(c) total distance to Chubut river (TreRawson).

*) Roberts overestimates (x 1.3) the distance between the encampment and the water-point in the same way (4 miles). The actual distance between Punta Cuevas and Laguna de Derbes is 2.8 miles. (Thomas Jones,Glan Camwy estimated it in 3 miles).

Table 3: Volume of stone removed from excavations.

Since each cave is a prism cut into a slanting surface, the resulting volume is a wedge. There is a rectangular back wall and two triangular lateral walls.

(A) surface (m²)

(B) volume (m3) (maximal height of the wedge: 1.70 m)

Hence, the 16 excavations could yield (7.6 x 16) m3 of stone but if the loss of material by breakage while cutting and moving the stones is evaluated in 25 % (arbitrarily), the volume of stone available to building is: (a) 7.6 m3 x 16 x 0.75 = 91.6 m3

Of the store built with this volume of stone, we simply know that it measured 5 m x 20 m145, that is to say 50 m of wall. We ignore the width of such walls as well as their height. It was a non-cemented wall raised with a very light stone (1.26 g/cm3), and the wind knocked it down in the first essay. When applying the nineteenth-century stone-building techniques (Rondelet's Ruler) 146, the width may be estimated as 0.65 m (for a 2 m high wall) (b) or 0.75 m (for a 2.5 m high wall) (c). The volume of stone employed in the construction should be, according to these conditions:

(b) 50 m x 0.65 m x 2 m = 65 m3

(c) 50 m x 0.75 m x 2.5 m = 94 m3

These figures (b & c) are congruent with de volume of stone available (a), especially if the number of unknown variables is considered. Obviously the stone available to build the store would decrease if a part were employed in closing the excavations themselves.

Existing hollows have an average surface of 9 m². One may suppose that the caves eroded by the sea had the same size. So, the 16 huts offered 144 m² of total covered surface. This allows to estimate that there was almost one m² available to each one of the colonists (if a standard 1.80 m x 0.55 m bed is considered). Clearly, the 61 children fit in less space. It is quite obvious, therefore, that there was not room enough to spare to lay up the goods of 26 families. Likewise, it is clearly stated that some people were logged in the store147. Present estimates of 9 people in each cave, are in good agreement with the figures given by Berwyn, that 8 people fit in each hut148.

1. Baur, John. 1954. The Welsh in Patagonia. The Hispanic American Historical Review. 34:468-498. Duke University Press.

2. Bowen, E. G. 1960. Welsh emigration overseas. Advancement of Science 17:265-270.

3. Williams, Glyn. 1991. The Welsh in Patagonia: The State and the Ethnic Community. Preface. University of Wales Press. Cardiff. 285 pp.

4. Bowen, E.G. op. cit.

5. Berresford-Ellis, Peter. 1984. The Celtic Revolution. 3rd.edn. Y Lolfa Ed. Talybont, 1993. 218 pp. Chap. VIII, p.177.

6. Meisen, Juan. 1992. Conclusiones de Juan Meisen Ebene sobre el desembarco del Mimosa y la fundación de Puerto Madryn. Centro de Estudios Históricos y Sociales. Transparencia n°16, p.12. Puerto Madryn, July.

7. Williams, G. op.cit. Preface.

8. Roberts, Edwyn. (MS1) Mordaith Edwyn K. Roberts o Gymru i Batagonia (unpublished). Parcially translated into Spanish in: Matthew H. Jones. 1981. Trelew un desafío patagónico. Tomo I. Ed. Golfo Nuevo. Puerto Madryn, p.121; and in: Biografía de Edwyn Roberts. Ellis Roberts. Eisteddfod Y Wladfa. Trelew 1986.

Roberts, E. (MS2) Hanes y Wladfa ar y Chupat (copybook). Manuscript n°759. North Wales University Library. Bangor.

9. Jones, Lewis. 1898. Y Wladfa Gymraeg yn Ne Amerig. Cwmni'r Wasg Genedlaethol Gymraeg. Caernarvon. 224 pp. Translated into Spanish as : La Colonia Galesa; Historia de una nueva Gales en Sudamérica. Ed. El Regional. Rawson, 1986. 237 pp. Chap.X, p.54.

10. Dumrauf, Clemente. 1991. Un Precursor de la colonización del Chubut: Henry Libanus Jones. Fundación Ameghino. Viedma. Doc. n° 10.

11. Humphreys, Lewis. 1910. Mordaith y Fintai Gyntaf. Dyddlyfr y diweddar Barch Lewis Humphreys gydag ychydig nodiadau rhwng cromfachau gan Richard J. Berwyn. Y Drafod n°590. Trelew, April 15.

12. Ap Iwan, Llwyd. 1889. Chronicle of a voyage to Valdes Peninsula in 1886. Y Celt. Y Bala, March 15.

13. Matthews, Abraham. 1894. Hanes y Wladfa Gymreig yn Patagonia. Aberdare. Translated into Spanish as : Crónica de la Colonia Galesa. (1954) 4th.edn. Ed. El Regional. Rawson, 1985. 160 pp.

14. Jones, Thomas. Glan Camwy. 1926. Hanes cychwyniad y Wladfa ym Mhatagonia. Y Drafod n° 1365 to 1383. Trelew, June 17 to October 22.

15. Humphreys, L. op.cit.

16. Jones, Joseph Seth. (MS, copy) Diary of the voyage aboard the "Mimosa". Museo Regional de Gaiman.

17. Evans, John Daniel. 1994. El Molinero. Ed.El Autor. Bs. Aires. 144 pp.

18. Destefani, Lucio. 1989. Historia marítima argentina. Tome VII. La Patagonia y Tierra del Fuego 1860-1870. Chap.XII, p.336. Departamento de Estudios Históricos Navales. Bs.Aires. 330 pp.

19. Díaz, Julio V. (1866). Informe sobre la Colonia de Gales en el río Chubut. In: Borgialli, C. 1934. Los fundadores de la Colonia Galesa. Argentina Austral 58:32-42

20. British Admiralty. (1871). Report from the Captain P.W. Dumistown of HMS "Cracker". In: Centeno, F. 1917. Orígenes de la Colonia Galense del Chubut. Revista de Derecho, Historia y Letras (19) 56:24-45. Bs.Aires.

21. British Admiralty. 1883. Report from the Captain H. Fairfax of HMS "Volage" (1876). Nautical Chart of New Gulf. Museo Regional de Gaiman.

22. Lasserre, Augusto. (1881) Plano de Puerto Roca. In: Dumrauf, C. 1993. El Ferrocarril Central del Chubut. Centro de Estudios Históricos y Sociales de Puerto Madryn. Doc.n°1. p.38.

23. Williams, Edward J. (1884). Plano de los terrenos concedidos a la Compañía Ferrocarril Central Chubut. In: Dumrauf, C. 1993. ibid. Doc.n°27. plate 1.

24. Phillips, Thomas B. 1962. The Valdean Expedition. (1885). Camwy n°3. Gaiman, January.

25. Ap Iwan, Ll. op.cit.

26. Mordaith y Fintai Gyntaf. Dyddlyfr y diweddar Barch Lewis Humphreys gydag ychydig nodiadau rhwng cromfachau gan Richard J. Berwyn. Y Drafod n°590. Trelew, April 15, 1910.

27. Semanario Golfo Nuevo. Pt. Madryn. n°878, December 9, 1931.

28. Berwyn, Ithel. (1935). Un trío de pioneers de la Patagonia. Argentina Austral. Selección 1929-1968. Tome 1:623-628. Bs. Aires, 1978.

29. Thomas, Evan. 1950. Los Galeses en el Chubut. In: Diario Esquel. Special Issue 25° Aniversary. Esquel. pp.65-67.

30. Davies, Horwell. 1960. De Gales a la Patagonia. Argentina Austral Selección 1929-1968. Tome 3:75-80. Bs. Aires, 1984.

31. Williams, Bryn. 1962. Y Wladfa. University of Wales Press. Cardiff.

32. Bowen, E. G. 1966. The Welsh Colony in Patagonia 1865-1885: A study in Historical Geography. Geographical Journal 132:16-31.

33. Williams, G. 1975. The desert and the dream. Wales University Press. Cardiff. 230 pp. Chap.IV, p.43.

34. Meisen, J. op.cit.

35. Jones, L. op.cit. Chap.X, p.62.

36. Roberts, E. MS1. op.cit

37. Roberts, E. MS2. op.cit.

38. Matthews, A. op.cit. Chap.V, p.22.

39. Williams, G. 1975. op.cit.

40. Jones, T. op.cit. n°1368, July 9.

41. Berwyn, R. op.cit.

42. Jones, Richard "Glyn Du". 1902. Recuerdos de los diez primeros años de la Colonia Chubut : La primera víctima del Desierto. Y Drafod n° 327. Trelew.

43. Jones, T. op.cit. n°1368, July 9.

44. Jones, Matthew H. 1981. Trelew un desafío patagónico. Tome I: 1886-1903. Golfo Nuevo Ed. Puerto Madryn. 318 pp. Chap.IX, p.128.

45. Humphreys, L. op.cit.

46. Jones, T. op.cit. n°1368, July 9.

47. Roberts, E. MS2. op.cit.

48. Jones, R. op.cit.

49. Berwyn, Richard. (MS). Register of births, marriages and deaths kept by RJB for the population of the Lower Chubut Valley during the years 1865-1875. Museo Regional de Gaiman.

50. Jones, T. op.cit. n°1371, July 30.

51. ibid. n°1381, October 8.

52. Evans, J. D. op.cit.

53. Jones, T. op.cit. n°1378, September 17.

54. Roberts, E. MS1 op.cit.

55. Jones, L. op.cit. Chap.X, p.62.

56. ibid. Chap.XI, p.73.

57. Evans, J. D. op.cit.

58. Matthews, A. op.cit. Chap. VII, p.28.

59. Lasserre (1881) in Dumrauf, C. 1993. op.cit.

60. Lambert, C. 1883. The voyage of the "Wanderer". Ed. Gerald Young. Macmillan & Co. London.

61. British Admiralty. 1883. op.cit.

62. Roberts, E. MS1. op.cit.

63. Matthews, A. op.cit. Chap.V, p.23.

64. Jones, T. op.cit. n°1368, July 9.

65. Díaz, J. op.cit.

66. British Admiralty (1871) op.cit.

67. Williams, E. J. (1884) op.cit.

68. ...The land rises gradually for about 4 miles or more, then there was a plain, as vast as the sea, over 10 miles and at last an entirely untilled hilly land, until the Chubut river, 45 miles away. Roberts, E. MS1. op.cit.

69. Ejército Argentino. 1967. Instituto Geográfico Militar. Hoja Topográfica de la República Argentina. Escala 1:100.000 Hojas n°4366-17/-18/-23 y -24. IGM. Bs. Aires.

70. Roberts, E. MS1. op.cit.

71. Meisen, J. op.cit.

72. The shore between Punta Cuevas and Punta Este is suitable ... to debarking with flat-bottomed boats. With easterly winds, it is possible to land a half-mile eastward from Punta Cuevas or immediately close to the rocks on the western side of the point. Armada Argentina. 1978. Servicio de Hidrografía Naval. Derrotero Argentino. Parte II H-202 7th.edn. Bs. Aires. p. 231.

It is an interesting fact to note here, that the day of the anchoring of the Mimosa, at about 1 p.m., there was a strong breeze from the sea, (easterly) at about 4 p.m. the Mimosa anchored in the port so expected. (Roberts, E. MS1.op.cit.). Apparently the ship could have cast anchor near Punta Cuevas and there was no hindrance in landing in the adjacent inlet.

73. These are, Lasserre's map (1881) and Edward William's railway map (1884)in Dumrauf, C. 1993. op. cit. and Evans, J. D. 1994. op.cit.

74. Gentile, E. 1994. Mapa Comercial y turístico de Puerto Madryn.

75. Phillips, T. B. op.cit.

76. Ap Iwan, Ll. op.cit.

77. Evans, J. D. op.cit.

78. Jones, L. op.cit. Chap.XI, p.73.

79. Roberts, E. MS1 and MS2.

80. Jones, T. op.cit. n°1372, August 6.

81. Evans, J. D. op.cit.

82. Berwyn, R. 1910. op.cit.

83. Jones, H. L. (1854) in Dumrauf, C. 1991. Un Precursor de la colonización del Chubut: Henry Libanus Jones. Ed. Fundación Ameghino. Viedma. 94 pp. Doc. n° 10. p.71.

84. Jones, L. op.cit. Chap.X, p.62.

85. Berwyn, R. 1910. op.cit.

86. Hughes, Glyn. 1978. Richard Jones Berwyn. El Regional n°317. p.31-36.

87. Ap Iwan, Ll. op.cit.

88. Roberts, E. MS1. op.cit.

89. Berwyn, R. 1910. op.cit.

90. Jones, T. op.cit. n°1368, July 9.

91. Monti, Alejandro. Coastal geologist. Centro Nacional Patagónico (CENPAT). P.Madryn. (pers. comm.)

92. Berwyn, R. 1867. Y Faner. September 7.

93. Jones, J. S. MS. op.cit.

94. Deodat, Leoncio. 1932. Noticias históricas de las guaridas galesas en Madryn. Golfo Nuevo n°906. Puerto Madryn, July 9.

95. Williams, B. op.cit.

96. Roberts, E. MS1 op.cit.

97. Jones, T. op.cit. n°1378, September 17.

98. McAdam, Jim. Environmental Scientist. The Queen's University. Belfast. Ulster. (pers. comm.)

99. Trespailhié, Orestes. 1932. Las cuevas-habitaciones. Golfo Nuevo n°886. P.Madryn, February 20.

100. Lewis, Margaret. 1964. Saskatchewan History (17)3:119. Notes and Correspondence.

101. El Chubut & Jornada. (newspapers). Trelew, September 26, 1995.

102. Gómez Otero, Julieta. Archaeologist. Centro Nacional Patagónico (CENPAT).P.Madryn. (pers. comm.)

103. Dahinten, Silvia. Antropologist. Centro Nacional Patagónico (CENPAT). P.Madryn. (pers. comm.)

104. Guerra, Pedro. Eng. in Forestry. Centro de Investigación y Extensión Forestal Andino-Patagónico (CIEFAP). Esquel. Chubut. (pers. comm.)

105. Encyclopaedia Britannica. 1972. Vol.17. William Benton Pub. p.1090, and Tinto, J. 1979. Utilización de los recursos forestales argentinos. IFONA. (National Institute of Forestry). Pub. n°41. Bs. Aires.

106. Jones, L. op.cit. Chap.X, p.62.

107. Berwyn, R. MS. op.cit.

108. Golfo Nuevo. 1931. op.cit.

109. Roberts, E. MS1. op.cit.

110. Jones, H. L. in Dumrauf, C. 1991. op.cit.

111. Roberts, E. MS2. op.cit.

112. Roberts, E. MS1. op.cit.

113. ibid.

114. Mac Donald, Elvey. Historian. Aberystwyth. (pers. comm.)

115. Roberts, E. MS1. op.cit.

116. Armada Argentina, op.cit.

117. Deodat, L. op.cit.

118. Williams, B. op.cit.

119. Three of these remnants were analyzed. Results indicated:

Ash, Poplar and Araucaria. (Guerra, P. Esquel, pers. comm.).

We know that Ash wood was brought aboard the Mimosa (Roberts, Tomos. North Wales University Library. Archives. Bangor, pers. comm.). Poplar, was the most available timber in Patagones in the half-nineteenth century. Likely, the planks brought by Edwyn Roberts and Lewis Jones were poplar wood. Finally, Araucaria, at a first sight a rather incongruent result, might prove that the caves were occupied until the late-1880s. Since then, the Andes were frequented and it is the region where Araucaria grows.

120. Berwyn, R. 1910. op.cit.

121. Jones, T. op.cit. n°1368, July 9.

122. Berwyn, R. 1867. op.cit.

123. Jones, T. op.cit. n°1368, July 9.

124. ibid.

125. Lambert, C. op.cit.

126. Trespailhié, O. op.cit.

127. Deodat, L. op.cit.

128. Jones, L. op.cit. Chap.X, p.64.

129. Jones, T. op.cit. n°1382, October 15.

130. Hughes, G. 1978.a. Lewis Jones. El Regional n°317. p.17-22. 131. Davies, William H. 1970. Interview with... September 1. Saskatchewan Archives Board. R-6078.

132. ibid.

133. Pendle, George. 1959: The Welsh in Patagonia. Wales 6:13-22.

134. Taylor, J. in Bowen, E. 1966. op.cit. Discussion.

135. Golfo Nuevo, 1931. op.cit.

136. British Admiralty. 1883. op.cit.

137. Borgialli, Carlos. 1934. Los fundadores de la Colonia Galesa. Argentina Austral 58:32-42

138. Amaya, Lorenzo. (1935). Fontana el territoriano. 3rd.edn. Junta de Estudios Históricos del Chubut. Cuadernos de Historia del Chubut n°5. Trelew, January 1969. Chap.IV, p.30.

139. Dennler, Jorge. 1938. Cuevas de la Patagonia. Revista Geográfica Americana (12) 74:357-360. Bs. Aires.

140. Gorráiz-Beloqui, R. 1954. Contemplación de la aventura de los primeros colonizadores del Chubut. Argentina Austral 277:16-20 and 278:10-14.

141. Jones, T. op.cit. n°1378, September 17.

142. Meisen, J. op.cit.

143. in Roberts, E. MS2 op.cit. and in Jones, T. op.cit. n°1368, July 9.

144. ...at that time were not houses to accommodate the people who were staying to build the railway...so caves were the only alternative. My father (engineer) and mother had to live in one of those caves where I was born. Margaret Lewis. op.cit.

145. Jones, T. op.cit. n°1371, July 30.

146. Devillez, A. 1882. Elements de constructions civiles. Vol. 1. Paris. 384 pp.

147. Jones, T. op. cit. n°1368, July 9.

148. Berwyn, R. 1867. op.cit.

Legends of Figures

Figure 1: a. Atlantic coast of Patagonia with settlements existing in 1865. b. Study site in the New Gulf-Chubut area.

Figure 2: Western extremity of New Gulf (New Bay). a. Madryn site. Main topographic features related to the location of the first Welsh settlement. b. Punta Cuevas. Past and present features referred in the study. (Parenthesis indicate out-of-date names). (1) Data from the 1876 British chart; (2) Data from the 1881 Argentine chart.

Figure 3: Western side of Punta Cuevas: The caves. (Dotted line indicates the high-tide level). a. Sectors of existing (A) and disappeared (B) excavations. b. Existing caves; the degree of destruction increases southward.